Front Cover
 Title Page
 Half Title
 Table of Contents
 Factors shaping training and learning...
 Training and learning activities...
 Relevance of training and...
 Quality of training and learni...
 Efficiency of training and...
 Effectiveness outcomes and impacts...
 Annex I
 Annex II
 Annex III
 Annex IV
 Annex V
 Annex VI
 Annex VII
 Annex VIII
 Annex IX
 Annex X
 Annex XI
 Annex XII
 Annex XIII
 Annex XIV
 Annex XV
 Annex XVI
 Annex XVII
 Annex XVIII
 Annex XIX
 Annex XX

Title: Evluation and impact of training in the CGIAR
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00077472/00001
 Material Information
Title: Evluation and impact of training in the CGIAR
Physical Description: Book
Language: English
Creator: Consultative Group on International Agricultural Research, Science Council
Publisher: International Maize and Wheat Improvement Center
Publication Date: 2006
Subject: Farming   ( lcsh )
Agriculture   ( lcsh )
Farm life   ( lcsh )
Funding: Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
 Record Information
Bibliographic ID: UF00077472
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.


This item has the following downloads:

Evaluation_and_Impact_of_Training ( PDF )

Table of Contents
    Front Cover
        Front cover
    Title Page
        Page i
        Page ii
        Page iii
        Page iv
        Page v
        Page vi
        Page vii
        Page viii
        Page ix
        Page x
        Page xi
    Half Title
        Half title
    Table of Contents
        Table of contents
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
    Factors shaping training and learning in the CGIAR
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
    Training and learning activities in the CGIAR
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
    Relevance of training and learning
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
    Quality of training and learning
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
    Efficiency of training and learning
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
    Effectiveness outcomes and impacts of training and learning
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
    Annex I
        Page A-1
        Page A-2
    Annex II
        Page A-3
        Page A-4
    Annex III
        Page A-5
        Page A-6
    Annex IV
        Page A-7
        Page A-8
    Annex V
        Page A-9
    Annex VI
        Page A-10
    Annex VII
        Page A-11
    Annex VIII
        Page A-12
    Annex IX
        Page A-13
        Page A-14
        Page A-15
    Annex X
        Page A-16
        Page A-17
        Page A-18
        Page A-19
        Page A-20
    Annex XI
        Page A-21
        Page A-22
        Page A-23
        Page A-24
        Page A-25
        Page A-26
        Page A-27
    Annex XII
        Page A-28
        Page A-29
        Page A-30
    Annex XIII
        Page A-31
        Page A-32
        Page A-33
        Page A-34
        Page A-35
        Page A-36
        Page A-37
        Page A-38
        Page A-39
        Page A-40
    Annex XIV
        Page A-41
        Page A-42
        Page A-43
        Page A-44
        Page A-45
        Page A-46
        Page A-47
        Page A-48
        Page A-49
        Page A-50
        Page A-51
        Page A-52
        Page A-53
        Page A-54
        Page A-55
    Annex XV
        Page A-56
        Page A-57
        Page A-58
        Page A-59
        Page A-60
        Page A-61
        Page A-62
        Page A-63
        Page A-64
        Page A-65
    Annex XVI
        Page A-66
        Page A-67
        Page A-68
        Page A-69
        Page A-70
        Page A-71
        Page A-72
        Page A-73
        Page A-74
        Page A-75
        Page A-76
    Annex XVII
        Page A-77
        Page A-78
        Page A-79
        Page A-80
        Page A-81
        Page A-82
        Page A-83
        Page A-84
    Annex XVIII
        Page A-85
        Page A-86
        Page A-87
        Page A-88
        Page A-89
        Page A-90
        Page A-91
        Page A-92
        Page A-93
        Page A-94
        Page A-95
        Page A-96
        Page A-97
        Page A-98
        Page A-99
        Page A-100
        Page A-101
        Page A-102
        Page A-103
    Annex XIX
        Page A-104
        Page A-105
        Page A-106
        Page A-107
        Page A-108
        Page A-109
        Page A-110
        Page A-111
        Page A-112
    Annex XX
        Page A-113
        Page A-114
        Page A-115
        Page A-116
        Page A-117
        Page A-118
        Page A-119
        Page A-120
        Page A-121
        Page A-122
        Page A-123
        Page A-124
Full Text

Consultative Group on International Agricultural Research

Evaluation and Impact of Training
in the CGIAR

JULY 2006



Evaluation and Impact of Training in the CGIAR

Review Panel:

Elliot Ster (Chair)

Lucia de Vaccaro
John Lynam

Panel Secretary: Sirkka Immonen


JULY 2006

Science Council Commentary on the
Evaluation and Impact of Training in the CGIAR

April 2006

The Science Council discussed the report on the Evaluation and Impact of Training in the
CGIAR at its 6th meeting held at WARDA in Cotonou, Benin after a videoconference
presentation by the Chairman of the Panel, Dr Elliot Stern. The Science Council conveys it's
thanks to the Panel Chair and the two members, Dr Lucia de Vaccaro and Dr John Lynam for
the commitment they have shown to the study over an extended period of time and for a
coherent and logical report that provides a strategic perspective to guide future training and
capacity strengthening activities of the CGIAR. The Panel assembled a vast amount of
information from 1990 to 2004 on past and current training activities in the Centers and,
although as it points out the databases at the Centers were surprisingly deficient and
variable, was able to synthesize them in a manner that enabled major trends and issues to be
distilled. This provided a useful setting for the field visits, case studies and surveys that the
Panel undertook.

The draft Panel report was shared with all CGIAR Centers and the comments received from
several Centers were taken into account by the Panel. Unfortunately because of the short
time available for finalizing the report not all comments were available for consideration by
the Panel in the final report or for discussion at SC 6. The Center comments have, however,
been further considered in preparing this commentary.

Panel's Findings

The findings and conclusions of the Panel appear logical and reasonably well founded, in
spite of the lack of comprehensive data and systematic analysis on which they are primarily
based. Commendably the Panel provides suitable caveats where biases and shortcomings in
methods are evident. The major findings of the Panel are as follows:

Relevance and Quality of Training

*The most important single factor that has affected the evolution of training in the CGIAR
over the past decade has probably been the increase in project funding and the reduction
in unrestricted funds available for training per se. As a consequence, this has lowered the
yield on the CGIAR's large investment in training and learning (currently about US$30
million annually; estimated to be about US$380 million for the 15 year period considered
in this study) because of (a) difficulties in building a critical mass of scientists and
multidisciplinary teams (b) difficulty in effectively funding higher degree studies when
projects are of 2-3 years duration; (c) since projects have shorter time horizons, the
training activities do not necessarily lead to greater relevance to the institutional or wider
needs of the trainees; (d) reduction in pedagogic support to Center research staff; and (e)

1 Based on the annual financial reports of the CGIAR Centers in Annex V of the report.

reduction in Centers' capacity to access, adapt, translate and disseminate existing
training materials.
* There has been an increase in "informal" and short course training linked to collaborative
research. Internationally recruited scientists spend on average 25 per cent of their time on
formal and informal training activities and this has increased over the past five years.
This may reflect the increasingly inherent role of training and mentoring in collaborative
research for capacity strengthening.
* Based on a number of indicators for groups and individuals, formal training quality has
been high. The most important determinant of trainee satisfaction is the extent to which
their new knowledge and skills were put to use. Unfortunately in many instances this did
not ensue. This emphasizes the need to ensure that candidates should be accepted only if
suitable post-training provisions are made or are likely. Improved candidate selection
procedures were considered by Center staff as one of the most important ways to
improve quality.
* There has been an increase in the proportion of shorter training periods and a decline in
longer duration training, both for group and individual trainees, with individual training
being somewhat stable but involving an increasing proportion of higher degree students
and women (40 per cent currently). There is a trend to much larger numbers from SSA
than from other regions.
* In some Centers there has been a marked increase in group training involving extension
officers and farmers.

Effectiveness of Training

* The effectiveness of CGIAR training as reflected by the perceptions of the persons
interviewed and surveyed has been quite high and has been as much determined by the
conditions of the NARS as by the relevance or quality of the training.
* There are a number of factors that have influenced training effectiveness: (a) changes in
the NARS, with some getting stronger and their staff becoming peers of the Center staff
and others getting weaker with different training needs; (b) donor priorities and funding
arrangements in the CGIAR; and (c) changes in technology, e.g. information technology
opening up the possibilities for virtual delivery of both training and training materials.
* The changes in funding sources available to training and consequent weakening of the
Training Units in the Centers in the past ten years has been accompanied by a trend
towards the decentralization of training away from headquarters to the regions and from
group training to informal on-the-job individual training in the context of collaborative
research projects. This has led to a loss of corporate knowledge and best practices, which
has made it difficult to maintain consistently high quality standards. Quality assurance
protocols for planning, managing and evaluating formal and informal training should be
specified and followed routinely. These measures are needed to ensure the system's
investment in training is used to good effect, and this requires enhanced training
resources and expertise in the Centers.
* Related to the increasing trend to project-related training, some countries, including
some of the poorest, have experienced a sharp reduction in training of all kinds. Hence
there seems to be no clear relationship between the extent of poverty in a country and
CGIAR training investment.

* Records kept are incomplete and inconsistent in many Centers and are not compatible
among the Centers in the System. This seems to be one consequence of the
decentralization of training and moving it increasingly to projects. In addition,
incentives for systematic record keeping and using data for planning seem to have been

The Evolution of Future Demand for Training

* The greatest future demand from NARS will be for capacity building through specialized
short courses and individual non-degree and higher degree training, instead of
generalized training; in this respect CGIAR could do more on e-leaming and support to
local universities.
* There is a need for improved coordination of training in the Centers along with enhanced
pedagogic expertise.
* The amount of training outside the deemed comparative advantage of the Centers
appears to be small, especially with individual training. Training in most Centers is
closely defined by their research programs. The Panel considers this a legitimate
definition of Centers' roles and they should not be expected to address the NARS' wider
training and capacity building needs. It points out though that there is a risk that such an
approach better meets the needs of stronger NARS at the expense of weaker ones and
hence that specific training needs assessments of the latter should be conducted.

Observations on Findings and Conclusions

The SC is pleased that the Panel implicitly validates the approach that has been taken in the
new System Priorities that training and other capacity building be closely linked to agreed
priorities and research collaboration between NARS and the Centers. This is in spite of a
perception in the report that TAC was not and the SC is not supportive of training and that
this has contributed to the decline in unrestricted funding allocated by Centers to training.
Certainly the SC agrees with the Panel that training of farmers and extension staff is best left
to others with a clear comparative/complementary advantage, with the Centers focusing on
scientist capacity strengthening with clear IPG attributes. It seems that this move by some
Centers to train farmers and extensionists has been partly motivated by an imperative to
focus training on "...downstream dissemination capacity as opposed to research capacity..".
The substantial increase in the former in recent years is confirmation of the move by a
number of Centers into the development arena, which has been criticized by the SC. In
addition, related to these trends are the possible moral hazards associated with
encouragement by some donors of the use of performance indicators such as training person-
days, and reinforces the SC view that the performance management system must measure
real outputs, outcomes and impact and thus create the appropriate incentives. The SC
encourages Centers to define clear training and capacity building targets within their
research projects.

The SC is not convinced that more systematic training needs assessments are required. By
the Panel's own assessment, the Centers have done a good job of identifying the capacity
strengthening needs of NARS within the context of trends towards increased consultations,
collaborative research projects and partnerships. The SC accepts however that in this process

the weaker NARS may have experienced that their training needs are increasingly unmet.
However in the SC's view, the Centers generally do not have a comparative advantage in
supplying all the training that the weaker NARS may need. The Centers can however
provide useful knowledge through e-based systems as an input for others to provide the
training. Additionally, some of the countries with stronger NARS also have the largest
numbers of poor people, require more formal scientist training and better capacity to use
research for addressing poverty.

The Panel did not discuss the role and achievements of the Centers in providing training
materials, and most notably did not mention the initiatives by the Centers in providing
global knowledge via e-systems (for example the Rice Knowledge Bank of IRRI, the global
training materials of IFPRI and from the erstwhile ISNAR).The report would also have
benefited from more details on how the Centers can contribute to and strengthen University-
based training in general, and in the context of the virtual university initiative in particular2.

The Panel seems to overlook the fact that the reason why there may not be quantitative data
on the increasingly important component of informal learning in the Centers may be because
it is indeed informal. This makes it more difficult to document and evaluate per se. While the
SC concurs with the Panel's criticisms of the poor state of documentation and evaluation of
formal training, their recommendation for more explicit monitoring of informal learning
would be more meaningful if they had provided some guidance on how the Centers might
go about documenting and evaluating informal training. Indeed it would have been helpful
had the Panel indicated what was a minimum data set for all types of training. The SC
concurs with the Panel's notion that better documentation of informal training where and
how it takes place would allow Centers better to incorporate informal learning objectives
into research activities and plan these opportunities for addressing capacity building needs.

The study earns high marks as a strategic review of training in the CGIAR. However, as the
Panel itself acknowledges, for various reasons it was not able to assess the impact of the
investments in training the system has made (currently some US$30 million annually) on the
goals of the CGIAR. This is disappointing and raises the question of the value and
desirability of undertaking a specific impact assessment of components of the program
where the databases might allow such a study. Some of the country case studies in the
Annexes to the main report would provide promising starting points. They cite assertions
and anecdotal information on impacts, although causalities and attributions are not verified
or documented for the most part. An important issue to be addressed in such a study would
be the extent to which training by the CGIAR generates private benefits in the form of
increased remuneration and advancement opportunities to the trainee, and what additional
international public good benefits accrue over and above these to the institutions and the
economy to where the trainee returns; and of course importantly to the poor. The high
attrition rates of trainees after they return to their home countries and the recognition by
NARS that investments in them can be lost to other institutions (see the Bolivia case in the
Annex, p. 32), suggest that a large portion of the impacts of training might be private and not
public goods. Of course the fact that countries might gain rather than individual NARIs in

2 GOAFU, Global Open Agriculture and Food University

such instances does not imply that CGIAR investment in training is not appropriate from a
NPG or IPG perspective. However these are researchable issues that deserve to be explored
further by SPIA and the SC.

The case studies reported on by the Panel did not include any "strong" NARS. To the extent
that training in the CGIAR has increasingly been research- and researcher-led, and most has
emphasized host countries as pointed out by the Panel, then maybe the study has not
adequately captured those NARS who have benefited most. The SC suggests this offers a
further possible rationale for a follow-up study of the impacts of CGIAR training on a
selective basis.

The Panel was concerned that most researchers thought there were few positive incentives
for them to be involved in training. This would seem to be inconsistent with the figure of 25
per cent of time scientists are currently spending on training and with the sense that this is
increasing. However due to the lack of comparable figures from other research institutions, it
is not possible to reflect on whether this time is appropriate for capacity strengthening by the
CGIAR system. The SC recognizes that some of the time spent on informal training activities
with graduate assistants or NARS colleagues during research projects also counts as research
time for the scientist, and in fact involves a leveraging of the researcher's time in such a way
that research progress is more rapid than if the graduate assistants or NARS colleagues were
not present. Thus, the SC recognizes this "double counting" as a potential win-win situation
for the trainees, scientists and Centers. The Panel made no attempt to separate these two
intertwined products. However the SC believes that both outputs (capacity strengthening
and research) and the subsequent outcomes are vital for the system. The SC will review the
performance measurement system to ensure that both outputs are captured and rewarded.

The Panel notes the poor quality of the reviews of training undertaken by the Centers, with
few conducted by outsiders and the focus being on outcomes rather than effectiveness,
efficiency or strategies. EPMRs also did not in general focus on evaluating training. The SC
will consider how the latter might be more effectively used to assess training strategies,
plans and impacts and encourages the Centers to commission more external reviews of
training using independent scientific peers and training experts so that EPMRs can be better
equipped to address training in future.

The SC notes the Panel's views on likely future demand trends from the NARS but was not
able to discern from the report how these were derived by the Panel. It will be important for
the Centers to assess these for themselves as they will undoubtedly vary depending on the
NARS concerned and the Center's programs.

Observations on Recommendations3

Notably absent from the recommendations is any that relates to comparative advantage of
the CGIAR vis-a-vis other sources of supply for training. This is a key issue and one that the

3 The recommendations are paraphrased here in italics.

SC believes must receive further consideration. Some recent EPMRs have also raised this

The SC has the following comments on the 13 recommendations.

CGIAR System

1. Formal and full recognition of training as an indispensable component of the CGIAR's activities,
both for NARS strengthening and as a contribution to execution of Centers' research. Following
this recommendation, at the investor's level, implies finding adequate resources.

The SC endorses the Panel's reaffirmation of the importance of NARS capacity strengthening
as an integral component, and not simply a by-product of the work and mission of the

2. The System should develop a uniform set of criteria and indicators of training outputs and
outcomes. An inter-Center focal group should develop such a set and present it for approval by all
stakeholder groups.

SC recognizes the inconsistencies and voids in information on training activities, outputs and
outcomes and the fact that this results in a lack of adequate information on which to plan for
the evolution of training in the System, to make it more effective and efficient in terms of the
mission and goals of the CGIAR and in terms of supporting the System's new set of
priorities. The SC endorses the formation of an inter-Center focal group with possible input
from the SC in terms of criteria and indicators for quality and relevance.

3. The System needs to come to grips with the issues associated with the increasing dominance of
short term, restricted funding and the System should make provision to overcome the associated

SC recognizes the problems of organizing training related to the increasing restricted vs.
unrestricted funding as part of a larger issue that needs to be addressed by the System's


1. NARS need to develop a clearer understanding of the areas of training in which the CGIAR has a
comparative advantage. These areas relate to the Centers' research agenda.

SC endorses this recommendation and suggests that Centers have a key role in clarifying
their comparative/complementary advantages and at the same time can make contributions
to the NARS through inputs related to identifying alternative sources of supply for non-
CGIAR priority NARS training needs. This recommendation also emphasizes the
importance of strengthening the "partnership" approach to training activities in recognition
of the fact that System research priorities have been and will continue to be informed by
NARS priorities.

2. NARS need to make a stronger effort to clearly articulate their research and training needs. This
can improve the effectiveness of cooperation with the CGIAR.

While the SC recognizes the importance of such articulation of needs, it also appreciates that
such specification often is difficult, particularly in the weaker NARS and/or where
conflicting interests exist and adequate mechanisms for coordination and collaboration are
missing. As with the CGIAR exercise leading to the new System priorities, the CGIAR can
work with NARS to improve specification of training and capacity strengthening needs.

3. NARS and Centers need to take greater care in selection of candidates for CGIAR training, to
ensure that candidates chosen have appropriate qualifications and post-training institutional
support and operational facilities.

SC agrees with this recommendation and recognizes that some Centers already have in place
fairly strict candidate selection procedures that could be shared more with NARS and among

4. An implied recommendation is that the Centers should reduce their involvement in direct training
of farmers and extension workers, except as an integral part of ongoing Center research.

SC agrees with this recommendation, which is part of the larger debate within the System on
the role of the CGIAR in production of IPGs and the optimum position for the CGIAR along
the R4D continuum in different circumstances.


1. Centers should adopt a strategic stance that involves:
Continuing to carry out training and promote learning compatible with their research priorities
and develop strategies to do so in ways that strengthen (and sustain) NARS capacities
Taking into account characteristics of successful outcomes in the System, including: longer term
commitment by Centers, longer term funding commitments, existence of local institutional
support and leadership, a mixture of formal and informal tuni,,lg ,i, ii,i activities; and other
Taking into account the need for special strategies for weaker, under-resourced NARS;
Taking into account the Panel's recommendation to give high priority to support for local
universities and establishment of partnerships.
SC endorses this recommendation, recognizing that some Centers already have
developed well-articulated strategic stances with regard to training and have
considered many of the factors that the Panel suggests are important. However
because training outcomes generally are not monitored, Centers are not learning from
both successful and unsuccessful outcomes. The SC believes that there is ample room
for greater inter-Center cooperation and collaboration in developing improved
strategies and training functions across the System, as well as good opportunity for
increased inter-Center collaboration in actual training activities, in the same way that
inter-Center collaboration in research is taking place, e.g., through Challenge
Programs and other inter-Center programs.

2. Centers should all develop appropriate quality assurance protocols to be applied at all stages in
both formal and informal training; and activities should be subjected systematically to appropriate
planning, monitoring and evaluation procedures, as in research.
SC endorses this recommendation, which relates to the need for improved and more
systematic information gathering and analysis procedures in the Centers. This
inconsistency in, and lack of adequate data and information is a particular weakness
that the Panel identified as a major one.

3. The Panel provides 1.'.,i, -;li' on how the Centers can improve the efficiency and effectiveness of
their training functions by taking advantage of opportunities for sharing experiences, best
practice, functions and activities among Centers, e.g., through such mechanisms as the ICT-KM
Initiative Online Resource Project.
SC endorses the suggestion to take greater advantage of inter-Center opportunities to
improve training and learning functions and activities of the Centers. In fact, the SC
recognizes that the Centers already are moving in this direction and thus endorses
such on-going activities.

4. Ensuring better coordination within and among Centers where this will enhance quality and
SC endorses this recommendation.

5. To better cater for the heterogeneity of NARS and exploit the advantages of ICT such as e-
learning, the Centers embrace the latter more explicitly.
SC endorses this recommendation but notes that the Panel has not commented on the
possible role of the Global Open University on Food and Agriculture in this context.
It therefore would welcome the views of the Alliance Executive on the scope for the
GOUFA to provide a vehicle for this.

6. Closer coordination and cooperation among the Centers in strategic planning of training,
assembly of data bases, development of courseware etc.
SC endorses this recommendation.

In conclusion the SC encourages the Alliance Executive to consider the value and desirability
of a System-wide Capacity Building Program to coordinate and share information among
Centers on the training related functions mentioned in these Center specific
recommendations, among other tasks.

Transmittal Letter

Dr Per Pinstrup-Andersen
Chair, Science Council
Consultative Group on International Agricultural Research
Division of Nutritional Sciences
Cornell University
305 Savage Hall
Ithaca, NY 14853-6301, USA

27th March 2006

Dear Dr Pinstrup-Andersen,

Re CGIAR Training Study

On behalf of the Panel charged with this study, I am pleased to submit our final report to the
Science Council of the CGIAR.

As you will be aware the Panel has engaged in an ambitious and wide-ranging review and
evaluation in order to identify the contributions and impacts of training and learning within
the CGIAR. We very much hope that our report will assist the Science Council in its future
deliberations. We also hope it will prove helpful to all those within the System and in the
NARS, whom we have met in the course of the study, and who are committed to further
strengthening research partnerships between the CGIAR and the NARS through training and

We would like to thank you and your colleagues including those on the interim Science
Council, who have offered wise counsel and shown us patience and courtesy throughout our
work. Whilst it might be thought invidious to pick out any for special thanks, we would
certainly wish to acknowledge the inputs received from Hans Gregersen, Jim Ryan and Ken
Fischer who had the responsibility to steer the study. They did so throughout helpfully whilst
showing proper respect for the Panel's independence. Finally I would like to acknowledge on
behalf of all of Panel members the generous contributions made by Sirkka Immonen of the
Science Council Secretariat. Her diplomacy, technical expertise and knowledge of the System
have been invaluable.

I look forward to hearing how the study and its recommendations are taken forward in due

Elliot Stern,
Panel Chair

Evaluation and Impact of Training in the CGIAR

MARCH 2006


SUMMARY AND RECOMMENDATIONS......................................................... 1
1 IN T R O D U C T IO N .............................................................................. .............................................. 9
1.1 Study objectives................................. ................. ................................. 9
1.2 Design and implementation choices and methods........................... .............. 9
1 .3 T h is rep o rt................................................................................................. .................... 1 1
2 FACTORS SHAPING TRAINING AND LEARNING IN THE CGIAR.................................... 13
2.1 CGIAR commitment to training and capacity strengthening.................................... 13
2.2 The changing context of CGIAR training.......................... ......................... 15
2.3 The scope of training and learning................................ ................................ 18
2 .4 C o n clu sio n s......................................................................................................................... 20
3 TRAINING AND LEARNING ACTIVITIES IN THE CGIAR.............................. .............. 23
3 .1 D ata co llectio n ........................................................................................... ................... 23
3.2 V o lu m e of train in g ....................................................................................... .............. .. 24
3 .3 T rain ee g en d er............................................................................................. ................. 29
3 .4 N atio n alities train ed .......................................................................................................... 29
3 .5 T rain in g th em es........................................................................................... ................. 32
3 .6 In fo rm al T rain in g .............................................................................................................. 35
3.7 Conclusions on data and data collection..................... .............................. 36
4 RELEVANCE OF TRAINING AND LEARNING........................ ..... ........................... 39
4.1 Defining the relevance of training and learning..................... ..... ............. 39
4.2 NARS capacity strengthening as a Center priority.......................... ................. 41
4.3 Criteria for judging relevance........................................................ 43
4.4 Priority setting in Centers and the NARS.............................. ........................ 45
4.5 Factors shaping NARS priority-setting....................... ............................... 46
4.6 NARS' perception of relevance........................ ........ ............................... 47
4.7 C onclusions......................................................................................... ............. 47
5 QUALITY OF TRAINING AND LEARNING........................ ...... ............................. 49
5 .1 D efin in g qu ality .......................... .............................................................. ................... 49
5.2 Methods of quality assurance in Centers...................... ....................... 50
5.3 Feedback from ex-trainees, partners and NARS................................. ................. 54
5.4 Conclusions......................................................................................... .............. 57
6 EFFICIENCY OF TRAINING AND LEARNING........................ ...... ........................... 59
6.1 Understandings of efficiency............................................................... 59
6.2 Deployment and targeting of resources............................ .......................... 60
6.3 Coordination and economies of scale............................... ............................ 64
6.4 Specialisation and comparative advantage.......................... ........................ 66
6 .5 C o n clu sio n s......................................................................................................................... 69
7.1 Understanding 'effectiveness', outcomes and impacts....................................... 71
7.2 Regional 'scenarios' and aggregate responses......................... ..................... 73
7.3 Partners, training and 'results'......................... ....... .................................. 77
7.4 Country overviews and case studies........................... .... ............................. 79
7 .5 C o n clu sio n s...................................................................................................... ............ 84
8 C O N C L U SIO N S ............................................................................................... ......................... 87
A C K N O W LED G EM EN TS................................................................................................................... 97
A N N E X E S ......................................................................................................................... ............... A -1


This study was commissioned by the interim Science Council (iSC) to review training within
the CGIAR as it contributes to capacity strengthening in the NARS. The purpose of the study
was to evaluate the quality and relevance, efficiency and effectiveness terms of intermediate
impacts in strengthening of the NARS and, to the extent possible, impacts in the CGIAR's
goals. The study was expected to provide recommendations to help Centers, Donors, the
NARS and the System to strengthen and plan their future activities in relation to training and
capacity strengthening.

Several issues related to the scope and methodology of the study influenced its design. These
issues are discussed in detail in the first Chapter, Introduction, and include i) defining
training and in particular distinguishing it from learning that takes place informally in
work-places and networks; ii) distinguishing training and learning effects from those of
Center research and other outputs; iii) country and regional focus; iv) defining capacity
strengthening; v) nature of CGIAR inputs and interventions and distinguishing these from
those of other actors; vi) how to judge 'impact'; and vii) scope of data collection.
Information for the study was obtained from: existing Center records and surveys carried
out by the panel of Center researchers; those in Centers responsible for training, trainees and
Center research partners.

The changes in CGIAR context that have influenced the way training has been conducted
and resources are discussed in the Chapter 2, Factors shaping training and learning in the
CGIAR. The factors that have had major consequences for the orientation and provision of
training across the CGIAR over the last 10 years include both those internal to the CGIAR
System and the broader global changes in NARS, agricultural technologies and funding.
Over this period, the NARS have become more differentiated; some gaining strength and
taking a role of equal partners, whilst others were more fragile and under-resourced some
even becoming weaker. Changes in funding and specifically the predominance of project
funding, has forced Centers to adjust the organisation and delivery of training which has
become increasingly decentralised to researchers. The role of 'training units' that coordinate
training services and plan training provision has diminished. On balance, the panel
considered that this trend has had a negative effect on NARS' institutional strengthening
and has curtailed Centers' ability to fully exploit the considerable investments made in
training and learning. New technologies and new public demands have shaped the training
agenda to include new kinds of skills in advanced technologies and social sciences. The
design of training, including new pedagogical approaches, communication technologies and
informal ways of learning, have influenced the way training is now delivered across the

The Panel collected data for 15 years (1990-2004) on group and individual training and its
analysis and conclusions are presented in Chapter 3, Training and learning activities in the
CGIAR. Data were available only for formal training and any quantification of informal
training and learning was based on surveys and interviews. The Panel observed
considerable deficiencies in the way training records had been collected and databases were
constructed, which seriously hindered their use for evaluation purposes or for planning by

Centers themselves. Among the most notable trends discernable, there seem to have been
increases in the numbers of group training events and numbers of participants. In some
Centers, there has been substantial expansion in group trainee numbers, due partly to
training involving farmers and extension workers associated with collaborative research
with extension services and post production research. The increase in numbers may also
reflect inclusion in records of more different types of events and better overall recording of,
for instance, regional training away from Center headquarters. A more stable pattern over
the years was observed for individual training. A high proportion of the trainees have come
from host countries of the Centers, and a less than clear relationship between intensity of
training and poverty levels was observed. Some individual countries, including some of the
poorest have experienced a sharp reduction in training of all kinds. The relatively high
proportion of developed country trainees was also notable.

The Panel analysed the data for themes to assess the comparative advantage of Centers as
training providers and concluded that only a small proportion of the volume of training (in
terms of trainee days) has been allocated to topics that are not within the Centers' research
capacity and mandate. The themes of Crop Production, Crop Protection and Breeding have
continuously been among the most common themes, while the themes of Social Science and
Biotechnology have gained in relative importance.

The relevance of training to strengthening NARS' capacity is discussed in Chapter 4,
Relevance of Training and Learning. The panel found that CGIAR Center training is broadly
relevant to the capacity needs of NARS. They concluded that it is appropriate to assess
training relevance within the context of the research agenda which centers share with the
NARS (i.e. as opposed to a broader definition of NARS training and capacity strengthening
needs). However, Centers are formally committed to capacity strengthening and many
researchers within Centers as well as those with some responsibility for training and learning
are evidently dedicated to helping NARS strengthen their research base. There were
perceptions among researchers that relevance may have been reinforced in recent years by
the decentralisation of training to Center researchers conducting collaborative research
projects. However, the formal commitments of Center managements was not always so
clearcut such that research relevance may not necessarily have led to institutional
strengthening. Furthermore, where under-resourced NARS were dependent on Center
support there might be at risk of distorting NARS research priorities and associated priorities
for training in order to access resources. CGIAR collaboration with other agencies with a
complementary but more development-orientated mandate is needed to address broader
NARS' capacity needs, which are particularly challenging in Sub-Saharan Africa.

The issues of quality are discussed in Chapter 5, Quality of training and learning. Perceptions
of training quality, gauged through trainee surveys were mostly very positive. Less positive
judgements were associated with limited opportunities to apply newly acquired knowledge
and skills. It is, however, difficult to extrapolate from past satisfaction ratings to present
conditions when researchers are more in charge of planning and conducting training. There
were limited quality assurance (QA) systems in place for training in some Centers. The
weakness of QA systems is due to the reduction of the capacity of training units or functions
and lack of pedagogic expertise among Center staff. Individual and informal training quality
is not addressed or monitored by any explicit mechanism. There are many examples of good

practice in place at Centers related to policies, recruitment and selection of trainees, course
guidelines, pedagogic support for researchers, collection of trainee feed-back and the use of
quality assessment to improve training. To make training quality a priority issue, Center
management, and indeed the CGIAR system, needs to communicate its support for training
emphasising the importance of quality and provide incentives and funding for quality

In Chapter 6, Efficiency of training and learning, the Panel concludes that the pre-requisites for
the efficient management and delivery of training and learning are not in place in most
Centers. Examples of good practice are unevenly distributed. The most important deficits are
inadequate pedagogic and coordination resources within most Centers and the absence of
systematic financial and monitoring data. However, it should be emphasised that the true
efficiency of training and learning is its contribution to the effectiveness and take-up of
research. The Centers were unable to provide detailed data on the investment in different
types of training and the trends over time. From System records and surveys, it was
concluded that the investment by the CGIAR in training and learning through formal and
informal means continues to be high. About 25% of researchers' time was estimated to be
spent on these activities. However, there is no consistent coordination, backstopping, advice
and support in all Centers for assuring the efficiency of training against the investments
made. Likewise, the coordination between Centers is a problem especially in Africa where
synergies could be achieved. There are instances where Centers have been efficient, for
example by adapting specific training 'products' into generalisable 'global' goods thus
achieving economies of scale in their production and use. However, it appears that due to
the lack of coordination within and between Centers the allocation of resources to training
that has taken place has not been always planned in the most strategic fashion. Closer
cooperation with NARS is required to ensure that trainees not only come with the necessary
pre-requisites prior to training but also have adequate possibilities of putting their training
to use afterwards. Centers visited were clearly aware of their particular 'niche' as providers
of training. These niches were consistently recognized by the NARS and in the opinion of
the Panel constitute areas of genuine comparative advantage. In general, Centers provide
training within their mandate "doing what they do best" although the Panel questioned
increases in volumes of 'farmer training' in some Centers and in some years. The Panel also
concluded that Centers should avoid covering resource shortages in NARS out of project
funds that cannot be sustained or select trainees without adequate preparation. To address
the broader capacity issues, coordination with other stakeholders, especially governments,
donors and universities is needed.

The Panel found no evidence to suggest that any single type(or types) of training were more
efficient than others. They concluded that Centers should continue to provide a mixture of
group and individual training activities, and achieve increases in efficiency mainly by fitting
these more closely to trainee and NARS needs.

The Panel's approach to evaluating outcomes and impacts from training and its analysis and
conclusions are presented in Chapter 7, Effectiveness: Outcomes and impacts of training and
learning. The Panel found strong and consistent evidence of the effectiveness of CGIAR
investments in training and learning. The case studies in seven countries across Latin-
America, Asia, and Sub-Saharan Africa confirmed that CGIAR training has led to impacts for

individuals and institutions. Many of the leaders of national research in agriculture are
Center graduates and the agricultural research agendas of NARI, government ministries and
other NARS partners have been shaped by Center inputs. In particular CGIAR centers have
contributed to the internationalisation of research linking even fragile NARS partners to
international scientific agendas.

Country studies and surveys of NARS partners confirmed the difficulty of separating out
training and learning effects from those of research and indeed germplasm distribution.
However, survey respondents' perceptions confirm that training was a significant
contributor to positive outcomes from research. Country studies also confirmed the growing
importance of informal training and learning alongside formal courses.

Contextual factors outside the control of the CGIAR limits the effectiveness of its
contributions to capacity strengthening. There are regional differences in sustaining and
using training and skills acquired with the Centers and institutional instability is an
important limiting factor, particularly in the poorer countries. This is illustrated by
'WASTAGE' rates among trainees in some countries. However the success and contribution
of CGIAR inputs have been striking even under the most adverse conditions, especially
when working with innovative local partners and committed donors. The sustainability of
the results of past investments in training and learning increases considerably when account
is taken of a broader set of 'results' that go beyond intentions and objectives. Projects have
left behind a large 'footprint' and many investments in training and learning have had
unintended but with hindsight foreseeable positive consequences for NARS.

The serious problems faced by countries where NARS are weak and where Centers in
isolation can only expect to have limited impacts, highlight the need for innovative
approaches to capacity strengthening. These will need to better integrate training and
learning with other capacity strengthening measures and coordinate the plans of more than
one Center together with those of other key stakeholders NARS partners, donors,
governments, and universities.


Recommendations can be variously directed to the CGIAR System, the NARS and Centers.

For the CGIAR System, the Panel recommends:
1. Training should be fully recognized as an indispensable component of the CGIAR's
activities, not only as a contribution to NARS institutional strengthening, but also as a
contribution to the execution and refinement of the Centers' research. At the investors'
level, full recognition implies finding, or helping to find, increased resources for basic
training support functions in order to optimise yields on the major investment currently
made and sustain the reach and effectiveness of collaborative research.
2. Given the investment of the System in training, simple but meaningful criteria and
indicators of training outputs and outcomes should be defined and used at the System
level, avoiding the current need to present the information in different formats to suit
different stakeholders. The definition of these indicators might best be undertaken by an
inter-Center focal group, but should be ratified and observed by all stakeholders.

3. The shortcomings of short-term project funding from the point of view of NARS
institutional strengthening must be recognized, and provision made to overcome them as
far as possible through integrated, longer term center-NARS-investor cooperation and
commitments. Training in association with research project funding may be putting the
weaker NARS at a disadvantage, and this situation should be revised periodically.

For NARS, the Panel recommends:
4. There should be a clearer understanding among NARS as to the areas of their training
needs which can be covered by the CGIAR. These refer to the areas of their research
agenda which they share with the Centers, and where priorities are set through mutual
5. In some cases, the absence of clear policy and articulated research/training needs on the
part of the NARS constitutes an obstacle to effective cooperation with the CGIAR. NARS
and Centers should work closely to improve this, possibly by more active intervention at
the highest policy level.
6. To safeguard the NARS' investment in training by the CGIAR, greater care must be taken
to select candidates with appropriate qualifications in coordination with the centers, and
to ensure adequate post-training support and operational facilities. For training to be
effective it needs to take place in the context adequate institutional support and where
necessary policy consistent.
7. For the CGIAR to support the NARS as effectively as possible, the Centers should not be
drawn beyond the limits of their distinctive competence as research institutions, into
activities which are the responsibility of national governments. In particular, their work
should be complemented by the necessary efforts to ensure downstream dissemination of
research products. In this context, the Panel would question the involvement of the
CGIAR in the direct training of farmers and extension workers except as an integral part
of ongoing Center research.

For the CGIAR Centers the Panel recommends:
8. CGIAR Centers should adopt a strategic stance with regard to the links and potential
benefits to NARS of the training and learning activities that they undertake. To this end
Centers should:
Continue to carry out training and promote learning compatible with their research
priorities and mandates and develop strategies to do so in ways that strengthens NARS
capacity. With regard to capacity building requirements which they cannot cover they
should cooperate and enable these to be met by other agencies and stakeholders
including international donors and national governments.

In developing their training strategies, take into account that cases with successful
outcomes encountered by this Panel often had in common: long term commitment by the
Centers; a long-term funding commitment; local institutional support and leadership; a
mixture of formal and informal training/learning activities, designed to fit specific needs;
the formation of multi-disciplinary teams and critical mass of scientists; a latent (or
explicit) demand for the technology in question that meets identified needs. These
'conditions' for success are likely to be valid in many NARS scenarios today.

Weaker, under-resourced NARS will need special strategies if poverty alleviation
objectives are to be met. Interventions at the highest policy level that will often also
involve informal learning opportunities (e.g. through policy dialogue) and an emphasis
on support for local universities through training and research partnerships may be the
options with best potential for long term impact. Close inter-Center cooperation and the
development of a common policy for capacity development including training should be
considered in such cases. The current distribution of trainee nationalities should be
revised at each Center to ensure it is justified on the basis of potential for poverty

In general, the Panel recommends giving high priority to support to local universities, as
probably the most sustainable contribution to capacity building through training. It
should also contribute directly to elevating the pre-training preparation levels of CGIAR
trainees. Various modes are already in practice, but the Center-north-south university
partnerships have particular merits. Partnerships with teaching institutions will also help
fill the Centers' gaps in pedagogic skills.

9. CGIAR Centers should ensure that formal and informal training and learning activities
should be systematically submitted to appropriate planning, monitoring and evaluation
procedures, as is research. To this end:
a) Quality Assurance protocols should be developed and applied systematically to all
stages in planning, managing and delivering training and learning, including needs
analysis and the routine specification of learning objectives in all projects.
b) In-country informal learning built in to projects should be supported by self-
evaluation guidelines that can be applied by project partners on a continuing self-
help basis.
c) Training quality should be systematically monitored and evaluated, routinely at the
immediate post-training stage. Long-term follow up studies of outcomes and impact
are only recommended strategically in samples of areas/projects. However if records
are well-kept and systematized across all Centers the present prohibitive costs of
follow-up would be dramatically reduced.
d) Training evaluations should be taken into account in staff performance ratings and
used to support the integration of training into research planning and decision

Important gains in efficiency are foreseen from ensuring that every Center has access to some
form of training and learning function and expertise however organised (the form will need
to vary to fit Center mandates and circumstances). In some circumstances these 'functions'
may be partly based within Centers and partly outside e.g. shared among Centers or at a
System level as with System's ICT-KM Initiatives Online Resource Project. These functions
a) Provide scientists at each Center with access to expert advice on suitable pedagogic
methods and delivery modes for training; retrieving, adapting and disseminating
existing training materials; and making materials widely available on-line.
b) Ensure the coordination of training activities across and between Centers where this
will bring benefits of quality and coherence, for example by systematising needs
analyses; facilitating inter-Center cooperation; implementing stricter candidate

selection procedures (see above); targeting national universities for training and as
partners in collaborative research projects; promoting learning alliances and center-
north-south institutional collaboration.
c) Given the heterogeneity of NARS, a variety of training themes, types and delivery
modes should continue to be provided, with emphasis on fitting them more carefully
to clients' needs, while making full use of ICTs and other contemporary methods. e-
learning for example can be a valuable complement within many kinds of training
and learning activities and alongside other forms of delivery face to face,
experiential etc. Specific e-courses can also be suitable for certain kinds of learners
and for certain kinds of content. The yield from the learning and training resources of
Centers will be better exploited in such ways.
d) Closer cooperation and coordination should be achieved in areas such as: strategic
planning, including regional/country strategies; the preparation, cataloguing and
delivery of materials; data base and financial recording system design to ensure a
minimum essential set across Centers in compatible formats; Quality Assurance
systems and related protocols; performance indicators; collaboration with other
sectors of the CGIAR related to capacity building (e.g. Information, Communications
groups), and exchange of best practices. A suitable inter-center mechanism (e.g. focal
point) should be set up, with funding, in order to achieve these objectives.


This section outlines the terms of reference and objectives of the study, gives some
background and contextual information to training in the CGIAR, indicates some of the main
design and implementation decisions made in the course of the study and outlines the main
sections of the report that follow.

1.1 Study objectives

This study was commissioned by the interim Science Council (iSC) to review training within
the CGIAR as it contributes to capacity strengthening in the NARS4. When commissioned,
the study was seen as part of a broader strategic priority for the iSC: the role of the CGIAR in
NARS strengthening.

The main objectives, as stated in the Terms of Reference (Annex I) are to evaluate:
* The relevance and quality of training activities carried out by the CGIAR;
* The efficiency and effectiveness of training; and
* To assess the intermediate impact of training in NARS capacity and, as far as possible,
the impact of training on the ultimate goals of the CGIAR.

The study was always intended to be forward looking as well as building on past and recent
experience. This was reflected in the expectation in the Terms of Reference that it would help
Centers, Donors, the NARS and the System to strengthen and plan their future activities in
relation to training and capacity strengthening. This future orientation was also emphasised
by the two Standing Panels having oversight of the study during its design stage.

Against this background, the Panel5 defined the overall aim of the study as follows:
To assess how far and in what ways the CGIAR System has provided and can best provide
training (based on scientific research) that strengthens NARS' capacity to undertake
collaborative scientific research to realize the goals of poverty alleviation, food security and
sustainable production.

1.2 Design and implementation choices and methods

A number of issues were identified in the course of designing and implementing this study
that have shaped its focus and outputs. The main design and implementation choices were:

* Issue: Defining training.
* Decision: To include a full range of formal and informal training/learning activities in
order to reflect the range of relevant activities that were encountered in preliminary

4 The term NARS has been interpreted throughout this report in the broad sense to include what is sometimes
labelled NARES and NARDS. The diversification of NARS and the active participation in national systems of
extension sector, NGOs, farmers' organizations and other development actors has been acknowledged in the
5 Biodata of the Panel members is given in Annex II.

investigations and pilot work. This includes learning which takes place in the course of
collaborative research, and networking when intended to develop and support training
and learning.

* Issue: Difficulty separating training and research and other Center inputs (e.g. germplasm
* Decision: To retain a focus on training and learning but not exclude activities that are
highly integrated with research and to try where possible to assess the value added or
contribution of training and learning whilst not expecting to attribute all results to

* Issue: Country and regional focus.
* Decision: To concentrate efforts in seven small to medium countries in three regions
(LAC, SSA and the Greater Mekong Basin within Asia) as these represented the likely
current and future locus of most CGIAR capacity development efforts and were
manageable within this study's available resources. A pre-requisite for inclusion was
that the country should have been a major recipient of CGIAR training, as indicated by
the data base compiled for this study. Those selected were: Bolivia, Cameroon, Ecuador,
Kenya, Malawi, Thailand, and Vietnam.

* Issue: Definition of capacity development.
* Decision: To define capacity development (consistent with current understandings) at
several levels in terms of individual capacities and skills; organisational capacities
resources and management; and inter-organisational coordination and networking.

* Issue: What constitutes CGIAR inputs and interventions?
* Decision: To recognize the importance of context. Accepting a broad definition of training
'interventions' (see above) underlines that CGIAR interventions occur in a context of
many actors which shape what is achieved and achievable.

* Issue: Focus of impact study element.
* Decision: To concentrate primarily on impact in relation to NARS' capacity and then,
where possible, on impacts for farmers and CGIAR goals.

* Issue: Scope of data collection.
* Decision: To gather data at several levels system-wide, centers, Country/NARS and
partners in order to cross check and be able to trace the factors that shaped outcomes
and impacts.

The methods and data sources for this study have included:
* Assembling a data-base on training types, volumes and trends from 1990-2004;
* Secondary sources such as EPMRs, impact studies and other reports and assessments of
the CGIAR;
* Case studies of 6 CGIAR Centers CIAT, CIP, ICRAF, IITA, ILRI, IRRI. These were
selected primarily on the grounds of their major contribution to training in the countries
chosen for field work, but also because their location allowed travel costs for the study as
a whole to be kept within the budget;

* Questionnaires to all researchers and training officers (or those responsible for training)
in all CGIAR Centers;
* Questionnaires to those who attended group training in 2003 and as many trainees as
possible who received individual training in the period 1993-2003;
* Questionnaires to partners for whom contact information was provided by Centers;
* Interviews and documentary analysis with the NARS at HQ and operational levels in 7
* Case studies in 7 countries of outcomes and impacts of training/learning including
collaborative research that incorporates training or education or informal learning;
* Follow-up or 'tracking work' with CGIAR Partners and Trainees in 7 countries to
ascertain the 'survival' of CGIAR trainees within the NARS;
* Feed-back from stakeholders on this report, at various stages of its preparation.

In estimating response rates to the questionnaires a number of caveats are in order.
Researcher questionnaires were distributed via Centers and although the Panel is reasonably
confident that it was sent to all on regular employment (circa 690) there may have been some
variation in some Centers. The numbers cited below for trainees and partners refers to
numbers distributed drawn from a much larger list. However the lists proved to be highly
inaccurate, with many misspellings, old postal addresses and other inaccuracies. It became
clear in the course of country visits that many to whom questionnaires were sent did not
receive them. Granted these caveats estimated response rates were as follows: Center
researchers 690 distributed via Centers, 338 received response rate 49%; Center training
officers and those with special responsibilities in that area, 40 distributed, 38 received -
response rate 95%; ex-trainees 2850 distributed 359 received response rate 12.6%; and
partners in collaborative research projects 2470 distributed (nominated by Centers), 148
received response rate 6%. The Panel concluded that the response rates was good for
CGIAR staff but low for partners and trainees as commonly found in studies of this kind
(see Annex III). This probably introduces a positive bias into the results, the magnitude of
which cannot be estimated, since those less interested in training or with negative
experiences would have been less likely to reply. The bias may have been particularly strong
in the case of the research partners, because they were named by the Centers and the less
successful and less persistent ones would probably not have been included. Throughout the
report, therefore, the Panel has been cautious about basing conclusions solely on evidence
from the questionnaires, and tried wherever possible to corroborate from various additional
sources the trends which they pointed up. Further analyses were conducted on the some of
the survey results to detect the significance of differences due to various sources of variation
(e.g. the effect of subject area on trainee satisfaction) using Chi-squared and other tests.
Further statistical analyses were undertaken of those who were critical or negative to
understand their responses. As has been found in other surveys, there is no reason to believe
that the 'negatives' that did respond are atypical of the negatives that did not.

1.3 This report

This report focuses on findings drawing on all the main data sources. It is organised into six
main Chapters. These cover:
* Factors 'shaping' training and learning in the CGIAR: The report begins with a description of
the factors shaping training arrangements, organisation and priorities in the CGIAR over

the last 10 years. This includes changes in the broader context, funding arrangements and
developing understandings about how training and learning can be supported in
different settings.
* Training and learning activities in the CGIAR: Available aggregate data are then presented
on trends in formal 'group' and 'individual' training. Estimates of the scale and
importance of informal training are also given, based on researchers' reports of the time
spent thereon, and on field study information.
* Relevance of training and learning: This section highlights what we are able to say about
relevance, understood to include priorities and priority setting processes at Center level.
This includes plan-making and consultation with NARS, as well as systematic feedback
from NARS and trainees. The section draws on evidence from questionnaire surveys,
case studies and country based fieldwork.
* Quality of training and learning: This section considers quality both in terms of the
processes likely to ensure quality and evidence that such processes are used. It also
draws on feedback obtained from ex-trainees as to their judgements of quality.
* Effi', oi y of training and learning: This considers how resources are deployed and how
training activities are organised and managed. It draws primarily on Center and country
visits conducted by the Panel, and on questionnaire survey results. Existing impact
studies are used as a secondary source of information.
* Outcomes and impacts of training and learning: This section reports on the effectiveness of
training. This includes intermediate 'impacts' of training and learning on NARES
capacity, discernable effects for agricultural systems and farmers and where possible
contributions to the CGIAR's own goals such as poverty reduction, food security and
sustainable production. It draws mainly on the survey questionnaires, country reports
and case studies conducted by the Panel, and refers briefly to existing training impact

The final chapter draws together Conclusions and Recommendations. Conclusions are also
highlighted at the end of each of the main report chapters. Supporting evidence from
surveys, case-studies and country reports are included in the Annexes.


This chapter briefly sets the scene reviewing the factors that shape training and learning in
the CGIAR. It describes:
* the CGIAR commitment to training and NARS capacity-building;
* how training is funded and organised;
* the institutional, funding and wider context within which training and learning is
delivered; and,
* the evolution and differentiation in how training and learning is understood in the

The chapter introduces material at a general level that is analysed and discussed in greater
detail in later chapters.

2.1 CGIAR commitment to training and capacity strengthening

System-level commitment
The CGIAR has a global commitment to strengthening National Agricultural Research
Systems. This is reflected in its stated objectives which have evolved in the course of this
study. When the study began these were stated as follows:

The CGIAR supports institution building and capacity building-
globally, regionally and nationally-to strengthen the evolving
international agricultural research community, and enhance the
professional development of agricultural scientists in developing

The latest version of these objectives as they relate to capacity building is stated in the New
Research Priorities of the Science Council6 as follows:

The CGIAR priorities maintain the focus of the system on research.
However, the conduct of international agricultural research,
combined with the provision of world-class opportunities for
capacity strengthening, is a comparative advantage of the CGIAR.
Enhancing capacity in developing countries has been a major
accomplishment of the CGIAR in the past. This approach will
continue through program-related opportunities and through
involving appropriate partnerships to enhance innovation and
learning. A,1,l;iltl,11, specific research on institutions is designed
to identify the best means for policies and institutions to support
new agricultural research and create pro-poor benefits.

This commitment is formally reflected in the mandates, objectives and activities of individual
Centers and in particular in their training and education activities. This is especially so as in

6 System Priorities for CGIAR Research 2005-2015, CGIAR Science Council, December 2005.

the CGIAR there tends to be a close identification of training and education with capacity

Linking Center research and capacity priorities
From Centers' own plans, objectives and other documentation, the primary purpose of
training activities is to enhance developing country organizations, mostly NARS, to be more
effective in independently and collaboratively conducting research for solving problems
primarily related to agriculture, environment and economy. The Centers focus their training
efforts globally and regionally depending on the mandate and focus of their research.
However Centers also emphasise the aim to train within their specific area of competence
and often the near term purpose is to improve capacity in that particular area of research and
activity. So for example IRRI has a general objective to 'generate and disseminate rice related
knowledge and technology of short and long term environmental, social, and economic
benefit and help enhance national rice research and extension systems' and sees training and
education as central to delivering that objective. Scientists are aware about how training
connects with their own research priorities: in the word of one, training is about 'helping
(this Center) implement our research that we think is important for the country and has
scientific value'. Balancing the needs of their own research and the capacity needs of NARS
is one of the challenges for Centers that this study will highlight.

As many researchers also acknowledge, the benefits of engagement with NARS is not one
way. Capacity building can variously create capacities to undertake research, give greater
focus to research and help in the formulation of new research agendas. This is discussed in
greater detail in Chapter 6, see especially 6.2.

Strengthening research capacity and potential 'partnerships'
The Centers aim at improving researchers' skills and knowledge about technologies and
methodologies, enabling some at least to become trainers themselves in the future.
Improving trainees' capacity to conduct further training is also highlighted by Centers.
ICARDA states their aim is the 'enhancement of researcher capacity to identify and
overcome constraints to production and understand the processes of technology transfer,
adoption and farmer decision making'. Stated objectives include the enhancement of the
development, dissemination, adoption and ultimately impact of technologies. One means for
achieving this is to establish collaborative partnerships for research and technology

World Agroforestry Center identifies institutional strengthening as one of
its four themes:
'We strengthen the capacity of institutions local, national and regional to
participate effectively in generating and applying innovations in agroforestry,
INRM, and environments for improved livelihoods.'

With regard to research systems and institutions it aims:
'to understand the bottlenecks faced by national institutions and to work out joint
strategies and programs to address them.'

Training at CIP

CIP's training program is a vehicle for interaction and collaboration with a
wide range of partners facilitating the achievement of the Center's
objectives. It is strongly linked with the research agenda and responds to
partners' needs for enhanced research skills and methods. It provides
effective mechanisms for the introduction of technologies to achieve
sustainable improvements in the productivity and utilization of CIP's
mandate crops, potato, sweet potato, Andean root and tuber crops, and in
the management of natural resources in the developing world.

The training program's aim is the creation of an international network of
highly capable research scientists able to conduct independent studies, to
offer skills training to others, and to collaborate effectively in the CIP global
community of interest.

ICRISAT has as an intermediate goal:

'Building partner power: R&D partners empowered through enhanced and
more relevant skills that include the ability to prioritize for impact, to
implement interventions and to predict trends.'

A strong incentive for Centers is to build partnership between the CGIAR Centers and
researchers and organizations, mainly in the developing countries. As one senior center
manager put it: 'training is an investment in cooperation'. This also leads to a related
purpose: facilitating partnership building between the organizations and researchers
receiving training. Training is seen as a two-way process that 'helps the Center streamline its
research priorities' (CIP). ISNAR specifically stated that the purpose of training is to
understand behaviour and attitudes of those who contribute to research alliances. In
addition to partnerships with developing countries, there are currently important efforts by
Centers to promote South-North (e.g. CIAT- Makerere University University of Florida)
and South-South partnerships (e.g. joint appointments with Southern universities such as
CIAT- University of Nairobi).

2.2 The changing context of CGIAR training

The differentiation of the NARS
Over the period under study the environment within which the Centers developed their
training strategies and resource commitments changed significantly.
* NARS in some developing countries significantly strengthened their agricultural research
capacity, and moved into newer areas such as molecular genetics and natural resource
management. Dependency on external research expertise and support gave way in these
NARS to stronger national capacity and nationally determined priorities. This is
exemplified in this study in the cases of Thailand and Vietnam.
* Poorer developing economies underwent structural adjustment programs during the
1990's that significantly constrained government spending, especially in the area of

agricultural research. These reductions in resources were accompanied by reductions in
agricultural research and in some cases the near-collapse of NARIs and public
universities. These problems were compounded in many of the same countries by the toll
of HIV/AIDS on agricultural research skills; and by the consequences of political conflicts
and civil war.
* Periods of political instability especially in parts of Latin America and Africa, shifts in
donor priorities (or in some cases capabilities given their own financial pressures) meant
that capacities of these countries significantly weakened during the period. This included
capacities within faculties of agriculture in the public universities.
* An important influence on the possibilities of CGIAR 'partnership' working worldwide,
was the entry into agricultural research in the 1990s of new classes of institutions -
mainly NGOs often with little research experience, and consequently, making new
demands of Centers for training. This was partly a matter of the changing role of the
State following on from structural adjustment but it was also the consequence of the
CGIAR, donors and NARS becoming more pre-occupied with 'impact' for poor farmers
and consumers.

The implications of these contextual changes for the CGIAR were a much more differentiated
NARI and NARS some where capacity had increased, some where it had diminished; that
had different needs for training in terms of sometimes more and sometimes less
sophisticated skills; and where capacity strengthening includes Universities, NGOs and
farmers organizations as well as NARI.

From core to project funding
One of the most potent 'shapers' of Center training over the period was
the shift in funding from core resources to project-based funding. Thus:

'The ratio of restricted funding to total funding rose to 55% in 2004 from
35% in 1995. Conversely, in 1995, unrestricted funding dropped from
approximately 65% of total funding in 1995 to 45% in 2004 due to the
high increase in restricted funding .....'
(Final Report, Task Force on Funding System Priorities, 2005)

The way that these system wide changes have impacted on particular Centers varies greatly.
However the effect has been to reduce 'unrestricted' funds to as little as 29% for IITA and
30% for World AgroForestry Center (ICRAF) and to maximum levels of 50% and 46% for
CIFOR and ILRI respectively. (See Annex IV on funding of CGIAR Centers).

At the same time there have been increases in overall resources available to CGIAR Centers
(according to the Task Force on Funding System Priorities, an increase of 32% between 2000-
2004) however most of this has been in restricted or project funds.

Detailed breakdowns of Center expenditure in terms of the deployment of core (unrestricted)
funds to training are difficult to obtain given the way budgets and costs are recorded.
However we were able to obtain figures for some Centers which demonstrate different
patterns of resource allocation and these are discussed in various parts of the report.

The organisation of training
Changes in volumes and categories of funding had large effects on how training was
organized, funded and implemented across the CGIAR. Most Centers found it difficult to
fund training as a stand alone activity from restricted project funding. Most of the training
funds were therefore incorporated into research project funding. However, this left little for
core support to training units, particularly when the limited core resources were utilized to
fund administration and longer term research areas such as genetic resources and breeding.
Many Centers during this period changed their training organisation and in effect
decentralised responsibility for training to research scientists relying on their ability to attract
funding for training within their research projects. At present, most Centers retain a Training
Units of some kind. Some Centers (e.g. IPGRI, ICRAF and IFPRI) have a capacity
strengthening as a project within the MTP portfolio and some (e.g. CIMMYT and IRRI) have
training within an MTP Project. However even Centers with Training Units and designated
capacity strengthening and training programs may have limited capacity. According to
survey data gathered from those responsible for Center training only 7 out of 15 Centers
have staff with any qualifications in training, pedagogy or adult education. (The
consequences of these organisational and capacity issues are considered in greater detail in
various parts of this report see especially Chapters on 'Efficiency' and 'Relevance'.)

Decentralization of training to researchers and research programs was often accompanied by
decentralisation of research and training to national and regional programs. For example:
* In the mid-1990's Centers sought to devolve group training, particularly the so-called
production courses, to national partners. Whilst this is seen as a response to resource
cutbacks by some it is also viewed as a positive guarantor of the relevance of training to
Center mandates by others. Devolution often involved training of trainers in a period of
declining national resources. This did not always lead to the hoped-for results, unless the
Centers themselves carried out the courses within the national programs. Many
'devolved' courses were taken back by Centers following initial difficulties.
* Survey results and Country and Center fieldwork have suggested that there has been a
significant increase in country based (rather than Headquarter based) training which has,
however, not been accompanied by the creation of new administrative systems to
monitor and manage what was being delivered. This study has found little or no
systematic information about country delivered training and learning a point that is
referred to throughout this report. It can even be argued that in-country training has not
increased as much as it would appear only that recording has improved. However
respondents to the survey of Training Officers or 'focal points' suggest that in 6 of the 13
Centers which provided information over 50% of their training now takes place outside
headquarters. This proportion has increased at 6 Centers, remained about the same in 5
and decreased in 2, during the last 5 years.
* In the 1990s there was also a trend across the CGIAR to create regional research
programs, particularly in Africa where the major portion of research funding was being
directed. This was intended to give Centers the potential to reach a wider cross-section
of clients. It had the consequence of shifting much of the training and capacity building
activities, particularly in the regional programs, to building what might be termed an
'impact pathway', that is the extension, farmer, and market capacities to have impact

with new technology. These trends shifted the focus of training towards extension-
workers and farmers in addition to scientists employed in NARI.

Emerging issues shaping training
There were many other external 'drivers' shaping Center training profiles in more particular
ways. For example Center based scientists cited:
* competition between developing countries;
* the biodiversity convention;
* the emergence of new technologies especially genomics;
* environmental pressures including drought and pesticides;
* producer-consumer market chains;
* the possibilities and potential of IT for training and learning dissemination, management
and delivery.

All the items of the above list create new demand for training and in some cases shape how
training was delivered.

2.3 The scope of training and learning

The word 'training' is generally understood as instruction or teaching within CGIAR
discourse. Such instruction or teaching may take place in courses (in 'groups') or
individually. However the system tends to downplay other learning opportunities that are
important even in an instructional setting e.g. interaction with fellow students in a course,
experiential learning in a field station or the relationship with supervisors in a graduate
degree program. (Chapter 3 has shown the importance of these activities.) There is certainly
little explicit acknowledgement of learning that takes place informally, through learning by
doing, work experience, learning in seminars and workshops, policy dialogue and in
research mentoring or in practitioner networks. These types of learning are not generally
monitored in CGIAR nor are they the subject of explicit learning management or quality
assurance methods. One indicator of this is that quantitative and administrative data on
informal learning is hard to find. This is despite the prevalence of many such learning
opportunities in diverse settings among CGIAR Centers. Adopting a broader perspective is
consistent with the findings of other studies of vocational training especially in
professional settings where training and human resource investments are increasingly
understood in terms of how and where people learn rather than in terms of what trainers

It became clear in the course of pilot work that many of the benefits of training in the CGIAR
derived from these broader expressions of 'learning'. The study has therefore consistently
sought to focus on how and more importantly, where learning occurs. It is for this reason
that the terms 'training and learning' are used extensively in this report. This emphasizes the
importance of learning that takes place outside of formal instruction and which requires a
shift in mind-set if issues of quality, relevance, efficiency and effectiveness are to be
adequately addressed.

Analytically and based on the case material available it is useful to distinguish between the
different 'learning strategies' adopted by Centers or more precisely its researchers and others

who are involved in training and learning activities. On occasions they may indeed be
'instructors' but at other times researchers pursue their learning objectives as managers of
networks or collaborative research or as mentors. At the same time there are different
learning modes the ways that those we call 'learners' and those we call 'teachers' interact.
Conveying technical content is very different from facilitating experiential learning or
facilitating peer learning. Learning or training strategies and different learning modes also
tend to take place in different settings and are likely to be appropriate for different learners
or trainees. The table below begins to unpack some of these distinctions. It is a framework
that has evolved iteratively beginning from a curiosity about how learning occurs within
and around what is called training in the CGIAR. However it was only during fieldwork and
interviews that the particular expressions of learning and its delivery became clearer.

Table 2.1 The learning process adopted by Centers

Learning/training Learning modes & settings Who learns Example
Instructional: The Transmissive/didactic: Usually the NARI Germplasm
Center knows and courses in specialised settings scientist management; biotech
the trainee needs to at (regional) HQ with techniques
learn experienced teachers
Learning manager: Mixture of didactic and The NARI scientist Plant breeding that
the Center manages experiential learning and NARES and combines a course
opportunities for learning by doing. Setting is to a limited extent, element and a period
learning more likely to be 'in-country' through research, on 'station' applying
the Center course knowledge;
scientists. Latter research assignments
lead in agenda designed or allocated
setting by Center to NARI
Mentor/advisor/seni Collaborative/peer learning Both the NARES 'Farmer participatory
or colleague: the through joint and Center the selection' collaborative
Center supports research/activities/projects, learning agenda is design and
learners mutual exchange between initiated by both customisation of 'tools'
Center/NARS; mentoring and methods or models
colleague exchange (both
individual 'visits' and
collective events seminars,
workshops). Technical advice
Network manager: Linking together diverse Limited or no Networks made up of
the Center brings research and development initiation by different
together related projects so as to help them CGIAR Center. projects/scientists in
leam from each others' Responsive or different countries to
experience/contexts and make dialogical which CGIAR scientists
explicit what they know. are attached
Meetings, workshops,
conferences as learning

The first column in the table concerns training and learning strategies. It progresses from the
simplest training setting where those who need to learn are 'instructed' through to more
facilitated and network-based strategies where there is less inequality between 'teacher' and
'learner'. The second column describes learning modes and settings. Learning modes progress
from what in pedagogics would be described as transmissive or didactic (within
instructional strategies) where teachers structure and deliver what they know, through to the
more experiential and collaborative modes of learning that take place in work settings and
collaborative networks. As this column also indicates these different modes are associated
with different settings. Transmission is common in classrooms but advisory missions and
joint seminars between Centers and NARS partners are more commonly associated with
collaborative learning and exchanges amongst peers. The third column focuses on who learns.
Here also it appears that there is a progression: from an instructional strategy where it is
mainly the 'trainee' who learns, though to the more reciprocal learning that happens when
Center based researchers working with NARS partners in networks and joint research
projects. The final column provides some examples of where these different configurations of
learning have been observed.

It is important to recognize that there can be no automatic assumption of 'progression' or
'development' moving down the columns in this table. NARS at early stages of their
development may remain dependent on instruction and imported skills and know-how for a
long time; and those NARS that have seen their development disrupted by political
instability as in Latin American case-study countries or by fiscal setbacks, disease and
conflicts as in Africa may move backwards from peer status and reciprocity to
instructional learning strategies and more dependent modes of learning. Nor can
generalisations be made even at the level of a single NARS. In some themes or disciplines a
NARS may well be relatively strong whilst in others it may lack capacity. It is also true that
when new techniques and methods emerge as has been the case recently in biotechnology
applications or post production/near to market methods -there is often a period when a
NARS reverts to instructional learning and training strategies or perhaps again works within
research projects designed by others.

What the table does suggest however is that NARS with greater capacity will tend to be
more autonomous and provide CGIAR Centers with research colleagues rather than trainees
and will learn collaboratively rather than through instruction. That is borne out by the
results of this study especially when comparing Latin American, Sub-Saharan and Asian
experience. Furthermore the table also suggests that there is a probable coherence across the
rows. It is difficult to deliver an instructional strategy except through some kind of classroom
(although this may come to be a virtual classroom in future as learning technologies and
associated skills improve). It is also difficult to imagine collaborative and peer learning
succeeding except in work based, joint research or network settings where there are
opportunities for learning by doing.

2.4 Conclusions

Overall the changes in CGIAR context driven sometimes by the CGIAR System and
sometimes by broader global changes in NARS, agricultural technologies and funding have
had major consequences for the orientation and provision of training across the CGIAR over

the last 10 years in particular although some of these developments have had a longer
gestation period.

Among the most important changes:
* NARS have become more differentiated previous Center 'trainees' in stronger NARS
have become colleagues and peers whilst some NARS have become more fragile and
under-resourced, their scientists still requiring basic training and support;
* Funding constraints have forced Centers to innovate in the organisation and delivery of
training in particular through the decentralisation to researchers and to country-based
* New technologies and new public policy concerns many of them connected with the
environment, international markets and poverty reduction have required the training of
successive cohorts of scientists in the technological and social science basics as well as in
more advanced techniques;
* Alongside these contextual changes there have been major changes in training and
learning with an increase in informal learning and the growing importance of
collaborative research, networks and peer learning alongside formal training courses,
whether for groups or individuals.

Not all of this is evident from aggregate data collected at a CGIAR System level and can even
be obscured by the way data is (or is not) collected. The next chapter draws together the data
that is available on formal training. In subsequent chapters when questionnaire results are
presented and discussed and NARS based case studies analysed, there will be more evidence
to support this broader typology of training and learning in the CGIAR.


This chapter presents the data collected from the CGIAR Centers on group and individual
training. The following aspects are described and discussed:
* data collection and problems associated with it;
* volume of training;
* gender and nationality of trainees;
* training themes;
* volume of informal training.

3.1 Data collection

Data collection began in 2001 during a desk study phase of the Training Study. During the
Main phase, Centers were asked to provide records for training up to 2004. Records for the
early 1990s in particular were difficult to obtain. Several Centers acknowledged that training
records were not systematically collected. Data for training outside the headquarters were
particularly patchy or altogether missing; and in some cases records had been compiled for
annual or other occasional reports and not into central databases. Given the variable
availability of specific data items and continuity of the data over the time period discussed in
the following, the data should not be regarded as providing accurate results of CGIAR
training, but rather as showing likely trends.

Data were originally collected on a large number of parameters. However, due to difficulties
in obtaining them, the Panel opted for a minimum set of parameters which include the
following annual information: number, length and theme of group training events; number,
gender and nationality of group training participants; number, gender, type and nationality
of individual trainees and the length and theme of study. To overcome the problems related
to gaps in the records, the Panel considered relative data and trends rather than the actual
figures when possible. In the trend analysis the Panel observed data in three periods: 1990-
92 (considerable gaps in the data) and two six year periods, 1993-1998 and 1999-2004 (good
data availability).

The largest gaps were in the records of nationality for group trainees, which were available
only for 37.4% of participants. Records on the type of trainee in group training were not
consistently recorded although such information was available for some Centers or was
occasionally to be found in the title event. The most complete data sets were obtained from
CIAT, CIMMYT, CIP, IITA, World Agroforestry Center (ICRAF), ICRISAT, IRRI [two data
sets: headquarters (HQ) and in-country (IC)], ICARDA and IPGRI. Data were also available
for the Systemwide program on Alternatives for Slash and Burn (ASB). IWMI did not
originally provide any data, but some records on individual training were available for 2003
and 2004. CIFOR also did not provide data, and it doesn't view itself as a training Center in a
conventional sense. CIFOR, however, provides capacity building both for individuals and
through organising occasional group events. ISNAR, which in 2004 became a program of
IFPRI, had training and capacity building as a major part of its agenda, but primary data on
training were not available in a form suitable for analysis.

3.2 Volume of training

Group training
In 1990-2004 there are records for about 90,000 people who attended group events that
Centers have included in the training data. Group trainees included 189 nationalities (see
section 3.4). Considering that the records for the early years and, in some cases, in-country
training are incomplete the total figure for group participants is certainly much higher than
the records show. However, the records for some Centers include very different type of
events, from formal group courses to conferences, meetings, field days and study tours,
some of which, particularly in the recent years have had a large number of participants as
discussed below. This makes an accurate estimation of the volume of group training
impossible and complicates meaningful interpretation of the kinds of training carried out,
and the types of trainees who were included. It is clear that group training of NARS7 staff
through courses, workshops and seminars is considerably less than the total reported here.

In providing data, Centers did not use similar definition of training (for instance ICARDA
and IPGRI data sets consist mainly of formal group training events) and this may be
reflected in the increasing vs. stable trends in Table 3.1. In general, the number of training
offerings remained at a similar level over 1990-2001 when the Centers organised on average
16 group training events annually. In 2002-2004 the average number of events was
considerably higher, about 32 events per year. This reflects a genuine rise in the number of
events carried out by ASB, CIMMYT, ICRAF and IRRI-in country (IRRI-IC) (Table 3.1), but
also in the latter case, more accurate recording. The trends with group training have not been
similar for all Centers.

IITA gradually brought group training to an end in 2001-03. At CIMMYT, ICRAF, and IRRI-
IC, group training has increased in terms of number of events in 1999-2004 compared with
the earlier years. ASB has also gradually increased the number of events since 1992 when
records started. At these Centers, as also at CIAT, IPGRI, IRRI-HQ, WARDA and WorldFish
the numbers of participants per event have increased in the last 6 years of the period
observed. At CIP and ICARDA there has been a downward trend in the number of training
events but the number of participants per event has remained similar. Only at ICRISAT there
seems to have been a downward trend in the number of participants per training event.

The summary trends for overall numbers of training participants in group events are
illustrated in Figure 3.1. The year 1994 is given as baseline because it is the first year with
records from all 14 Centers and the ASB program.

SNARS is here considered to include NARI, relevant government departments and institutions, universities,
NGOs and the private sector. The sectors include agriculture, forestry and fisheries.

Table 3.1 Changes in group training events and participant numbers 1990-2004

1990-1992 1993-1998 1999-2004
Events (annual average)
Increasing trend currently
ASB 10 21
CIAT 12 5 13
CIMMYT 9 14 41
IFPRI .. 10 14
ICRAF 8 8 47
IRRI-in country 8 23 54
WorldFish Center 26 39
Stable or decreasing trend currently
CIP 46 47 40
ICARDA 41 39 31
ICRISAT 12 12 10
IITA 15 15 4
ILRI 11 4 4
IPGRI 5 22 17
IRRI-HQ 18 11 15
WARDA 4 6 6
Participants/event (annual average)
Increasing trend currently
ASB .9 28
CIAT 11 16 26
CIMMYT 20 17 26
ICRAF 22 23 30
ILRI 15 12 18
IPGRI 14 11 18
IRRI-HQ 16 19 28
IRRI-IC 29 22 26
WARDA 20 25 30
WorldFish Center 20 31
Stable or decreasing trend currently
CIP 21 25 24
ICARDA 14 15 16
ICRISAT 11 10 6
IFPRI .. 21 22
IITA 17 16 18

In the recent years changes can be observed in some Center's training that are difficult to
interpret as the increase may be due to a number of factors. At CIAT, CIMMYT, ICRAF,
IRRI-IC, WorldFish Center and ASB total numbers of participants in group training have
gone up in recent years and were on average 5.5 times higher in 1999-2004 than in 1993-1998.
The sharpest rise was observed at ICRAF, ASB, IRRI-IC and WorldFish. In 2003-04, ICRAF
trained over 4000 group participants annually, compared with an average of 150 in 1990-2001
(data for 2002 missing). ASB trained 700-1800 participants annually in 2001, 2003 and 2004,
compared to an average 113 over the previous 10 years. IRRI-IC events involved 2300 to 8400

participants annually in 2002-2004, compared to 380 on average in 1990-2000 (data for 2001
missing) and, according to IRRI, the increase is due to systematic collection of records in
recent years. WorldFish trained 1200-1500 participants annually in 2000-2002 compared to
330 on average in 1993-1999 (data for 2003 and 2004 not available).

Figure 3.1 Relative change in the number of group training participants

9 / \
8 average of ASB, CIAT, CIMMYT, ICRAF,
SIRRI-HQ, IRRI-IC and WorldFish
o 4
3^ / A

~1 ^- -- ----"-^---^"-"

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
1994= 100% (1)

Some available records were omitted from the analysis because they represented what
appears as exceptional kind of activity and involved very large numbers of participants. In
addition to the IRRI-IC data considered here, the PETRRA8 network involved about 24000
participants in 2001-2003 in some 420 events. In 1994, WorldFish involved some 1100 farmers
in a "farmers' rally" and IRRI's in-country training event of one day on integrated pest
management involved 1440 participants. For CIMMYT, parallel recording in 2003 showed
that group training targeted to NARS participants included 58 events and 1918 participants9,
while a more comprehensive set of records covering a diversity of events and including, for
instance farming family training, contained 141 events with 9600 participants.

These figures reflect the same phenomenon discussed above of including in training records
wider range of events with larger numbers of participants from broader circles of
stakeholders than in previous years. Events geared towards farmers and extension staff on
hand and program, regional and international meetings on the other hand may have become
more frequent, or at least more frequently recorded. It could be assumed that events
involving very high numbers of participants were shorter than others, but records on event
length are not consistently available in these cases.

The sharp changes in trends in the recent years appear to reflect the inclusion in training
records of a wider range of events with larger numbers of participants from broader circles

8 Poverty Elimination through Rice Research Assistance
9 Included in this analysis

of stakeholders than in previous years, particularly those geared to farmers and extension
workers. An example taken from ICRAF shows that 29 of 171 events in 2003 involved 50
participants or more, and that farmers were identified as the participants in 10 of these,
accounting for a total of 1300 trainees, but on average these events lasted less than 2 days. In
2004, one single training event, "Introductory agroforestry, nursery management and aspects
of HIV/AIDS relationships with agroforestry" accounted for 555 participants. In WorldFish
Center's training data the peaks in 2000-2002 cannot be explained by increase in farmer
training. Rather the records show a high proportion of workshops and meetings, which
characteristically may have involved more participants than courses. For the earlier years,
such detail on the nature of the events was not available.

Some of these increases, or possibly the more comprehensive recording, may have been
triggered by the performance indicators used by the World Bank in 200310, which included
trainee days as one indicator and which were used for funding decision on a small part of
the World Bank's total allocation to the CGIAR. In any case, records of farmer and extension
events and program, regional and international meetings involving very large numbers of
participants influence the general data and make interpretation of trends difficult when
differentiation of different kinds of training is impossible or cumbersome.

Figure 3.2 Long training events as % of total number of events



Z 60 average for CIAT, CIMMYT, IITA, IPGRI and IRRI-HQ

40 '- -----.

20 average for CIP, ICARDA, ICRISAT and WorldFisb.
S..------. ,,, .
". A" .*

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Length of group training events
Data on the length of group events were available for 98% of the records of 10 Centers for
years 1993-2004, which were included in the analyses. Such data were available only for
some events or not at all for IFPRI, ILRI, IRRI-IC and ASB. In the analysis, events longer
than 30 days were considered long, and events of 10 days or less were considered short. In
the first half of the 1990s several Centers (CIMMYT, IRRI, ICRISAT and IPGRI in particular)
were offering long courses, which in 1999-2004 have accounted for only 10-20% of group

10 In 2004 and 2005 a CGIAR tailor made performance measurement system was introduced and volume of
NARS training was not longer an indicator.

events (Figure 3.2). At the same short group training events (<10 days) have become
relatively more common in most Centers; at CIAT, CIMMYT, IITA (before group training
stopped) and IPGRI. In the commodity Centers this trend may reflect the decline in long
term breeding and production training.

Individual training
In 1990-2004 the CGIAR Centers trained about 13,000 individuals. Records for at least some
of the years included in the study were available for 14 Centers. WorldFish did not have
usable records for individual training. Records from CIFOR and IWMI were very limited and
from WARDA covered only some of the parameters considered. Comparing the periods
1993-1998 and 1999-2004, the annual numbers have not changed with about 960 individuals
per year. In the earlier years fewer individuals were trained, which may reflect gaps in
records. Centers where comparison of the two periods shows more than 20% increase in
average annual training of individuals comparing the two periods include CIP, CIFOR,
ICRAF and ICRISAT. At CIMMYT, IITA and ILRI individual training has dropped more
than 20% from 1993-98 to 1999-2004.

The training records for individuals include long term on-the-job and degree training and
short term orientation and specialization training. Centers have classified individual training
in varying ways. A standard" that was introduced by the IARC/NARS Training Group in
early 1990s (database updated till 1996) has not been followed by other Centers except ILRI
that was the host of the database. Furthermore, in some cases, depending on the status of the
individuals, Centers included them in a visitor database, rather than in the training records.

Individual training has ranged from very short duration to several years. The length of is
clearly correlated to the type of training. The shortest duration, < 10 days stay, has increased
among non-degree trainees. Training of 2 years of longer has decreased among degree
trainees. More than 50% of the degree students for whom data were available, spent more
than 1 year at the Center, but from 1990-92 the proportion of those spending more than two
years at the Center has diminished.

Table 3.2 Changes in length of individual training

Length of stay 1990-92 1993-98 1999-2004
% of
% non-degree % degree % non-degree %d non-degree % of degree
trrainees trainees trainees trainees trainees

<10 days 6.5 2.1 9.7 0.3 14.1 0.6
>10 days < 30 days 33.7 0.6 25.3 1.2 24.6 1.4
1-6 months 45.7 12.8 44.3 17.7 43.3 19.6
6-12 months 8.2 13.4 13.2 15.1 12.0 25.2
1-2 years 5.3 18.9 5.2 17.7 4.5 17.4
over 2 years 0.7 52.1 2.3 48.1 1.4 35.8

11 Graduate fellow, research fellow, senior research fellow, student associate, technical associate, visiting scientist.

Relatively complete records on degree and non-degree training were available for 5 Centers,
and for 10 Centers data were available with gaps. The analysis of the 5 Centers' data show
that the relative number of degree students has increased from about 40% of trainees in 1990
to about 60% in 2003.

3.3 Trainee gender

Data on gender of the participants in group training events were available for 8 Centers for
2001-2004. In the period 1990-2004 the proportion of women increased from 17.1% to 20.7%.
IPGRI (30%), CIAT (26.9%), IRRI (26.4%) and IFPRI (25.1) have trained relatively more
women than the other Centers observed, while at ICARDA the proportion of women in
group training is relatively low (15.7%; data for 2001-2004).

Among individual trainees (records available for 89%) the proportion of women has been
considerably higher than among group participants and has increased from about 30% in
1993-98 to about 40% in 1999-2004. CIFOR, CIP and ICARDA have had the highest
proportion of women (45-50 %), while at CIMMYT and WARDA female students have been
less than 20% of individuals. There has been fluctuation from year to year, but in general the
proportion of women has increased or remained the same in all Centers and at CIAT, CIP,
ICRISAT about 50% of individual trainees were women in 2004 (at IWMI the ratio was also
nearly equal at 43% women).

3.4 Nationalities trained

Nationality information was available for group trainees from 10 Centers12 covering about
37% of all group trainee records and 59% of participants of these 10 Centers, and for 95% of
the individual trainee records from 13 Centers. For some Centers individual records in
general were available only for a few years (WARDA, CIFOR, IWMI). Overall, Centers have
trained nationals from 194 countries. The distribution of nationalities by region and Center in
individual training is shown in Table 3.3.

For group training the data on nationalities were too limited to permit meaningful
conclusions. Particularly the absence of in-country training records in many cases renders
the nationality information less useful, as it is likely that in-country training reaches different
nationalities in different proportions compared with headquarters events13. The records from
CIP, ICARDA and IRRI, where the volumes of group training were highest, dominated. The
data suggest that CIAT, CIP, ICRISAT, and WARDA have trained predominantly host region
nationals and also at ICARDA, IITA and ILRI host national were the largest group also less
than 20% of trainees. In CIAT's case the extent of regional training in Africa for instance, in
unknown. CIP's training records show more global reach in its group training than with
other Centers: Only about 62% of the group training participants were from Latin America.

13 IPGRI is an exception as it has only in-country training while records are centrally collected.

The second most common region for CIP's group training was Asia (19%) and about 12% of
CIP's group trainees came from SSA.

Table 3.3 Distribution of nationalities of individual trainees by region and Center14

Asia and Latin America Sub-Saharan Developed
Pacific and Caribbean Africa CWANA countries
Center n % % % % %

CIAT 1608 2.7 77.9 3.0 0.3 16.0
CIFOR 132 24.2 11.4 13.6 0.0 50.8
CIMMYT 1962 29.1 29.1 24.6 7.5 9.7
CIP 1669 6.0 81.2 5.5 1.6 5.6
ICARDA 1681 2.0 0.0 8.5 85.1 4.4
ICRAF 627 14.1 11 56.4 0.3 28.1
ICRISAT 1736 61.2 1.3 23.4 4.3 9.7
IFPRI 189 14.3 3.2 63.5 1.1 18.0
IITA 837 0.5 1.0 86.1 0.6 11.8
ILRI 767 1.2 0.7 82.9 0.4 14.9
IPGRI 518 20.8 23.4 18.3 10.1 23.4
IRRI-HQ 1114 81.7 0.7 3.1 2.8 11.7
IWMI 38 34.2 0.0 36.8 0.0 28.9
WARDA 73 1.0 0.6 83.5 0.3 14.6

Total 12951 23.3 26.2 24.9 13.8 11.9

In terms of which Centers provided most group training in specific regions, the
that IRRI, IITA, ICARDA and CIP have been the most prominent providers in
CWANA and LAC, respectively. CIP also trained considerable numbers
participants in Asia (16%) and SSA (15%).

data show
Asia, SSA,
of group

Among individuals trained, nationals from LAC, SSA and Asian countries have been trained
in approximately equal numbers (23-26%). It is noteworthy that the CGIAR Centers have
trained nearly as high a number of individuals from developed countries as from the
CWANA region. The proportion was highest at CIFOR, but in terms of numbers, CIAT led
with some 250 developed country trainees, CIMMYT and ICRAF trained about 190 each and
ICRISAT about 170.

The five most common nationalities for both group and individual trainees are listed in
Table 3.4 for all Centers for which any records were available. The percentage of trainees
from host countries is also shown.

CIMMYT and IFPRI: data for 1993-2004
CIFOR: data for 1995-2004
IWMI, WARDA: data for 2002-2004

Host country nationals account for a large proportion of Centers' group trainees although
these results are likely to be influenced by incomplete data, particularly for in-country
training. Individual records, however also show the predominance of host country nationals,
which for CIAT, ICARDA, ICRISAT, IITA and ILRI accounted for 30-50% of trainees.

Judging by the data for individual trainees, training in the Asian, CWANA and Latin
America regions has concentrated on one or two nationalities, namely India, Syria and Peru
and Colombia, respectively. In SSA three countries, Kenya, Ethiopia and Nigeria account for
45% of individual trainees. USA, Netherlands, Germany and France account for 50% of the
individual trainees from the developed countries.

Table 3.4 Predominant nationalities, including host country* of group and individual trainees

Group training
Colombia (66.1%), Ecuador,
Nicaragua, Venezuela, Peru

Peru (69.2%), China, Colombia,
Bolivia, Uganda
Syria (14.9%), Egypt, Iran, Afganistan,

India (40.4%), Myanmar, Bangladesh,
Vietnam, Malawi
Nigeria (16.7%), Ghana, Kenya,
Uganda, Mozambique
Ethiopia (19.0%), Kenya (13.8%),
Tanzania, Uganda, Nigeria
Kenya, Malawi, Bangladesh
Philippines, Bolivia, Malaysia, Mexico,
Philippines (22.0%), Cambodia, China,
Bangladesh, India

Individual training
Colombia (47.1%), Brazil, Ecuador,
Venezuela, Germany
Indonesia (25.4%), France, USA, UK,
China, Kenya, Mexico (7.8%), India,
Peru (25.4), Ecuador, China, Kenya,
Syria (41.3%), Iran, Ethiopia, Jordan,
Kenya (17.3%), Indonesia, Rwanda,
Netherlands, Uganda
India (46.8%), Germany, Sri Lanka,
Sudan, Vietnam
Nigeria (36.3%), Cameroon, Ghana,
Benin, Belgium
Kenya (26.1%), Ethiopia (19.7%),
Nigeria, Uganda, Germany
Uganda, Malawi, China
Colombia, Kenya, Peru, Ecuador,
Vietnam, India, Philippines (10.9%),
China, Bangladesh
Sri Lanka (15.8%)

WARDA Cote d'Ivoire (52.0%), Guinea, Ghana,
Mali, Burkina Faso
* Percentage of host country national given in brackets













The countries chosen for visits and country study were also among those that had received a
relatively high volume of individual and group training from more than one Center15.

The relative changes within each region are shown in detail in VI. The training of Indian
nationals has increased and was in 1999-2004 nearly 50% of all Asians. Training of Chinese
and Indonesians has also increased, while training of Vietnamese and Philippine nationals
has decreased. In Latin America the relative increased has been highest in Colombia, while
most others have decreased, including trainees from the countries chosen for case studies,
Ecuador and Bolivia. In Sub-Saharan Africa the training of different nationals has remained
at similar levels with a slight increase of Kenyans and a slight decreased of Ethiopians. In
CWANA training of Syrians has decreased in relative terms in 1993-1998 and 1999-2004 as
compared to 1990-1992 while training of Iranians increased.

All Centers have been training individuals in Asia and SSA. In terms of which Centers were
involved in training in each region, individual trainees from Asia were trained by ICRISAT
(34.4%), IRRI (29.7%) and CIMMYT (19.2%). The contribution of the other Centers ranged
from 0.1% to 4.9%. In SSA WARDA and ILRI both trained 20.4% of the individual trainees,
followed by IITA (16.9%), CIMMYT (13.3%), ICRISAT (9.5%) and ICRAF (8.8%). In Latin
America, most individual training was been done by CIAT (37.9%) and CIP (36.2%). Except
for CIMMYT (18.5) and IPGRI (4.7%) the contribution from other Centers was 1% or less. In
CWANA ICARDA trained the vast majority of individual students (77.8%) followed by
CIMMYT (8.8%), IPGRI (5.7%) and ICRISAT (4.1%). All Centers trained developed country
nationals, CIAT (14.7%), CIMMYT (11.6%) and ICRAF (10.6%) being the biggest contributors.

3.5 Training themes

The training themes were analysed for group and individual training on the basis of the
course, thesis or job title provided in Center records. Training was classified in 12 general
theme categories: Agroforestry, Breeding, Biotechnology, Crop Production, Crop Protection,
Genetic Resources, Livestock, Methods, Natural Resource Management (NRM), Post-harvest,
Seed and Social Science. All training topics not specific to any particular area of research
were classified under Methods. In individual training Methods accounted for a very small
proportion (see section on Themes in Individual Training below).

Theme information was available for the majority of group events and individual training.
Because both the group training events and the individual study periods were of highly
variable length, the volume of training in the different themes was analysed as trainee days
and as number of participants16. Total trainee days reflects the actual volume of training
more accurately than the number of events or participants, while the latter is a better
reflection of the breadth or coverage of the audience trained. The results of the group and

15 In Asia: Vietnam (4th most common nationality considering all records), Thailand (7th); Latin America: Bolivia
(3rd), Ecuador (4th); in SSA: Kenya (1), Cameroon (8th), Malawi (9th).
16 Trainee days data for themes were available for 54% of the group and 76% of individual training; participants
data for themes was available for 81% of the group and 78% of individual training. For group training the trainee
days could not be calculated for IRRI and IPRGI. For individuals, the data on length were very limited from
CIMMYT and ICARDA. Livestock is probably underrepresented, because records for ILRI seemed to be missing.

individual training are given in Tables 3.5 and 3.6, respectively. Data are presented in overall
percentages for each theme for the period 1990-2004 as a whole, and then ranked in order of
relative importance for three periods: 1990-1992 when records tended to be erratic, and in
1993-1998 and 1999-2004 when records were judged to be relatively complete. Details of
themes by Center are given in Annex VII.

Analysis of the data by length of training gave similar results for group and individual
training, but analysis by number of participants showed different themes as the most
common ones, as illustrated in Table 3.5 for group training and Table 3.6 for individual

Themes in group training
As the data in Table 3.5 show, even in the absence of IRRI data17, there has been a clear
predominance of Crop Production and Breeding in each of the three time periods in terms of
trainee days, although the relative numbers of participants fell in recent years and also the
volume of training in Breeding fell. Methods was important throughout, and ranked highest
overall in numbers of people trained, accounting for 16.2% of the total training 1990-2004.
The main change was the relative increases in terms of volume in Social Science (from rank 9
to 6 to 2) and Livestock (from rank 12 to 10 to 5) and relative decrease in Crop Protection
(from 5 to 4 to 9). The change in livestock is partly explained by missing data on course
length in the early years and by the events having been relatively long.

The most important changes over time in the coverage of people trained were the relative
increases in Seed, Social Sciences and NRM, with decreases in Crop Production, Breeding
and Crop Protection. Agroforestry became the second ranking theme in numbers of people
trained in 1999-2004, due to the vastly increased training of ICRAF. In some cases the
breadth of coverage was not reflected in the amount of time (i.e. trainee days) dedicated to
the theme. This indicates that the nature of training may be different depending on the
themes; training of Breeding, Biotechnology and Livestock involves more often long study
periods for relatively few people with an aim at in-depth competence in the theme, while
training of Agroforestry, NRM and Methods may have been more orientational, or aimed at
enhancement of a particular skill or aspect of the theme. Genetic Resources, then, is a theme
where the target audience is smaller than for the other themes.

A breakdown of the Methods category, shown in Annex VIII, indicates that Statistics/Data
Management accounted for the highest numbers of participants and trainee days, especially
if added to Experimental Design, which was classified separately. Together these sub-themes
explained about 30% of the Methods category in terms of trainee days and numbers of
participants over the whole period.

The theme Methods could be considered least associated with the research activities of the
Centers. It has, however, remained a common theme accounting for over 11% of all group
training in 1999-2004. The prevalence of this theme does not seem to reflect the shift of
training function and funding to research programs as it has remained near the top among

17 Trainee days could not be calculated

group training themes as judged by participant numbers and even in terms of group training
volume. In volume, the emphasis has been in methods, such as Research
Management/Process, Experimental Design and Statistics/Data Management, where the
Centers may have particular expertise and relevant orientation due to their research agenda.
The Centers may be the sole providers also in themes such as Information Technology,
Scientific Writing and Training & Education, which are among common Methods taught,
even if these training themes may be completely removed from the Center's research focus.

Table 3.5 Relative importance"8 of different themes in group training, in terms of trainee days (td)
and numbers of participants (p)


% td




% p ranking ranking ranking ranking ranking ranking
(td) (p) (td) (p) (td) (p)

Crop Production


Social Science

4 1

5.8 2

9.2 9


5 2 8 4 7

8 6 7 2 5

2 3 1

3 1

Crop Protection



6.4 8.5 5

4 4 3 9 8

6.1 3.1 10 12 5 10 7 10

6.0 1.2 12 11 10 12 5 13

5.9 12.7

4.6 8.7 4

3.2 3.8 7

6 7 2 6 3

8 5 8 6

9 9 9 11 11

Genetic Resources

2.7 6.5 11 10 11 6 10 9


2.0 9.4 6

7 12 11 12 2


0.7 1.5 13 13 13 13 13 12

Themes in Individual Training
In individual training Crop Protection, NRM and Breeding were outstanding in importance,
with little relative variation over the three time periods. In 1993-3004 Biotechnology ranked
third in terms of participants. The main changes were shown in the decrease in the relative
importance of Crop Production and Livestock. In contrast to the picture shown for group

18 Importance here refers to prevalence over the study period, and it is recognized that while themes may be of
equal importance the target audiences are not equally large for each theme.





training, Methods was of only moderate importance for individuals, especially in terms of
trainee days (3.2% of total trainee days).

Table 3.6 Relative importance of different themes in individual training, in terms of trainee days
(td) and numbers of participants (p)19

1990-2004 1990-1992 1993-1998 1999-2004

% td % p ranking ranking ranking ranking ranking ranking
(td) (p) (td) (p) (td) (p)

Crop Protection 18.7 17.3 2 1 2 1 1 1

NRM 17.3 12.5 3 4 1 4 3 4

Breeding 15.6 14.8 1 2 3 2 2 2

Biotechnology 11.1 12.5 6 8 6 3 4 3

Social Science 9.5 8.3 5 7 4 7 5 5

Crop Production 7.9 6.8 4 3 5 8 8 8

Agroforestry 6.0 3.4 9 12 8 9 6 9

Genetic Resources 5.8 7.3 8 10 7 5 7 7

Methods 3.1 8.2 10 5 10 6 9 6

Livestock 2.1 3.2 7 6 9 11 13 11

Post-harvest 1.5 2.0 12 11 11 12 10 12

Other 0.7 0.7 13 13 12 13 12 13

Seed 0.7 2.9 11 9 13 10 11 10

3.6 Informal Training

Informal training and learning has not been documented traditionally in the CGIAR, and this
report appears to be the first that has attempted to quantify its importance. As will be shown
in Chapter 6, researchers estimate that they spend an average of 12% of their total time on
this, which is about the same as on formal training activities (13%). To gain some insight into
what informal learning opportunities have arisen in the course of a collaborative research
project, an example is shown in the Ecuador Case Study 1, which describes Center staff
leadership and advisory roles, as well as visits to and from the Center for purposes other
than formal training. Taken together, the activities described suggest an extremely important

19 Data on length were very limited from CIMMYT and ICARDA that have the highest individual trainee
numbers. The volume of breeding, crop production and protection, and NRM, that have been frequent themes
with those two Centers, may therefore have been even higher than shown here.

learning contribution through leadership, advice and mentoring. It is significant that one of
the most consistent features of the Country visits was the importance trainees and partners
attached to the informal learning which takes place through, for example "learning from
colleagues on the job" or the long-term working relationships which have frequently
developed between Center staff and trainees. Testimony to this effect is provided in the case
studies (e.g. Bolivia, Cases 2, 4). Given the importance of this activity in terms of staff time,
and its perceived value to the trainees, it is inconsistent that there are apparently no
processes in place in the Centers to plan, document, monitor or evaluate it.

3.7 Conclusions on data and data collection

Conclusions on data systems
Data bases have not been kept systematically by all Centers. Some were discontinued during
the 1990's, presumably in association with the reduction of core funds to training, and
although others have been introduced recently, there is still no minimum essential data set
recorded routinely across the CGIAR Centers, or even within most individual Centers.
Consequently, basic information required for decision making on training within the CGIAR
system is lacking. One of the most significant gaps is meaningful information on who has
been trained, and their functions in the overall system. For example, it would be useful to
have a breakdown between policy makers, researchers, extension workers and farmers. The
records available at present have been collected for a particular purpose, such as annual or
project reports, and records contributing to the analysis in this study were for some Centers
obtained from many different sources. Commonly, fields in a database have not been filled.
Spelling mistakes and entries in variable formats (e.g. dates) can make sorting and querying
impossible. Lack of information on the costs of training also reflect a disconnection between
financial planning and reporting and program planning and reporting.

Overall these shortcomings seem to indicate a lack of appreciation of the benefits of
systematic record keeping, lack of communication between database managers and those
organising the training events, or entry of data after an event when details are no longer
available. There is a clear need to define a minimum data set for use across the Centers, with
simple but useful classifications of key items (such as trainee type) which will permit easy
sorting and meaningful interpretation of the results in future. These should be agreed upon
by stakeholders so that improvised requests for information in different formats are avoided.
Implementation of such classifications will need to be backed up by systems capable of
delivering information with consistency and accuracy.

The current state of data-gathering and monitoring systems with regard to training and
learning in Centers also seems to reflect a lack of incentives to do this well and a perception
that this is not an activity valued by the CGIAR as a whole.

Conclusions on available data
Among the notable trends in the results, there seem to be increases in the numbers of group
training events and numbers of participants in about half of the Centers, some of which have
showed a massive expansion in group trainee numbers, due partly to training farmers and
extension workers. A more stable pattern over the years is shown for individual training.
The information on nationalities shows a high proportion of host country trainees at most

Centers, and a less than clear relationship between intensity of training and poverty levels.
However, the latter may reflect a relatively lower number of suitable candidates from the
poorer countries, rather than a lack of intention to support them on the part of the Centers.
But the fact is that some individual countries, including some of the poorest (e.g. in LAC, see
Country Studies) have experienced a sharp reduction in training of all kinds. The relatively
high proportion of developed country trainees (12%) is notable. It may be partly due to
donor preferences and availability of suitable scholarships to support the trainee, as opposed
to Center policy, but appears to have reached levels which merit revision in some Centers.
With respect to training themes, one of the most controversial aspects refers to those which
are often considered outside the Centers' comparative advantage. The present results
suggest that these in fact correspond to a small proportion of total trainee days, especially in
the case of individuals.

With respect to training themes, the results show distinct trends over time in their relative
importance for group and individual training, although the traditionally predominant
themes in both cases remained fairly stable. Thus, the rise in relative importance of themes
such as Social Science (group) or Biotechnology (individual) was not at the expense of drastic
declines in the older subject areas such as Crop Production (group) or Crop Protection
(individual). One of the most controversial aspects refers to the subject areas which are often
considered outside the Centers' comparative advantage. The present results suggest that
these in fact constitute a small proportion of total trainee days, especially in the case of


This chapter assesses the relevance of training and learning to strengthening NARS capacity.
It begins by discussing how relevance and capacity are understood in the CGIAR and more
widely; reviews the evidence collected in the course of this study as to the commitment of
Centers to capacity strengthening; the perception of relevance by the NARS; considers some
of the factors that appear to be shaping NARS prioritizing and which constrain what Centers
are able to achieve; and finally draws overall conclusions and suggests measures that the
CGIAR Centers might adopt to further improve the relevance of their training and learning
activities to NARS strengthening.

4.1 Defining the relevance of training and learning

The Panel defined the relevance of training in terms of 'its applicability to strengthening
NARS capacity to undertake collaborative scientific research to realize the goals of poverty
alleviation, food security and sustainable production'. Consonant with the global mission of
the CGIAR, training activities should also meet the 'international public goods criterion'
(Inception Report, 2004).

Implicit in this definition are assumptions regarding:
* The role and contribution of training and learning in capacity strengthening;
* The nature of 'capacity' itself; and
* The goals being pursued and to which ends capacity is deployed.

Thus capacity is viewed in terms of its contribution to NARS being able to undertake
agricultural research; and links are made to the broader goals which NARS indubitably share
with the CGIAR in relation to hunger, poverty and environmental sustainability. In the
CGIAR where training and learning is nowadays mainly decentralised to researchers and
closely integrated with Centers own research strategies and mandates many interconnections
need to be taken into account. A simple model would then link the four elements of
training/learning, research strategy, capacity and goals as shown in Figure 4.1.

Figure 4.1 Model of training relevance

Research Training/
Priorities Learning NARS Goals
Mandate Capacity

In this representation training and learning related activities, in the context of Center
research priorities is directed at strengthening NARS capacity, which then allows the NARS
to pursue the shared goals. Relevance is a process of delivery (large arrows) and alignment,
as indicated by the smaller, feedback arrows. Alignment refers to a matching process that
requires information, gathering, prioritisation and mutual adjustment. Whether or not the
contribution of Center training and learning outputs is relevant to capacity strengthening of
the NARS, depends partly on judgements about what is delivered but also on the robustness
of the mechanisms in place to decide on priorities. The next section of this chapter therefore
considers what Centers see themselves as delivering by way of capacity strengthening and
then assesses processes of alignment the decisions made intended to ensure that training is
consistent with the needs of NARS.

What is delivered through the means of training and learning to strengthen NARS capacity
depends on how capacity is conceived. In the wider literature on institutional capacity
strengthening (see for example: Capacity Development, UNDP Technical Advisory Paper 2
1997, Horton, Douglas et al Evaluating Capacity Development ISNAR, IDRC, CTA 2003), it is
common to conceive of capacity at three different levels:
* individual capacity and skills;
* organisational capacity, including management arrangements; and,
* inter-institutional capacity, including networking.

All of the above are embedded in an 'enabling environment'. The panel has considered all
three levels of capacity the individual, organisational and inter-organisational in questions
asked in questionnaires and the checklists for NARS fieldwork and case studies. The wider
'enabling environment' has also been taken into account in national overviews and in
comparing training and learning results at a regional level in Africa, Latin America and Asia.

For CGIAR Centers, training activities nowadays derive from and are usually integrated
with Center mandates and research priorities. Decentralisation to researchers and research
programs is the norm. However there are also the goals of other parties to consider. In the
first place NARS have goals which may be better or less well articulated and which may
or may not overlap in their entirety with Center mandates and priorities. Furthermore there
are the main goals of the CGIAR sustainable agriculture, poverty reduction and food
security the salience of which will differ across different Centers and different parts of their
research and training portfolios. Thus whilst in both figures below there is a high level of
coherence between Center mandates and training and learning inputs, there is a greater
consistency between NARS priorities and needs and Center training and learning inputs in
Figure B than in Figure A. It is also assumed in Figures A and B that CGIAR goals are always
broader than those of any one Center.

The implications of the different configurations represented in Figures A and B are discussed

Figure A

Figure B

4.2 NARS capacity strengthening as a Center priority

Questionnaire surveys and case studies of NARS and Centers all confirm that NARS
strengthening including through training and learning is a priority in the CGIAR. This is
reflected in policy and strategic plans and backed up by the views of researchers. When
asked to rate the importance of 'training and structured learning activities aimed at NARS
strengthening' for their Center over 85% of researchers responding to questionnaires rated
this as 'important' or 'very important'. When asked to assess the importance of capacity
strengthening for their own research, 68% rated it as 'important' or 'very important'.
Outcomes of training and learning were also reported by researchers responding to
questionnaires to include all the different aspects of capacity strengthening from 'trainees'
career opportunities are improved' to 'new capacities and skills embedded in NARS' and
'new networks created'.

Similarly, according to Training Officers (or focal points) institutional strengthening is
judged overwhelmingly (by 96%, 29 out of 30 respondents) the 'most important output of

Center training'; and the two most important criteria in deciding the kinds of training that
takes place are 'Center's mandate' and 'Demand from NARS'. This is reinforced by the
trainee survey data: most trainees were employed by NARS when beginning their training
and most of them were encouraged to participate in training and learning by their

Perhaps the main inconsistency between the Centers' declared commitment to training and
what happens in practice, is evidenced by the generalised reduction of unrestricted funds to
training during the 1990's, which affected the training units, services and support. Given the
budget reductions, Centers chose to channel funds out of training to sustain other activities.
Some of these, such as gene bank maintenance which was also severely underfunded, are far
less resilient than training to budget fluctuations, and in this sense the decisions were
justified. But the experience suggests that the Centers' commitment to training has in fact
been strong up to the limit where the continuation of training related activities puts at risk
other vital long-term functions which are even more essential to the research mandates. A
second inconsistency is that a majority of researchers (55%) report that there are few positive
incentives in Centers to become involved in training and learning activities. This was
explained and elaborated in open-ended comments:

Performance evaluation is stacked grossly in favour of research, very little to gain by doing training -
No clear institutional messages that training is important No or very little funds Institutional
culture views capacity building as soft and not important

Upper management never mentions training. General feeling is that training is no longer considered a
priority. That functions as a potent disincentive.

We, at (Center X), consider training a very important function. Past decisions to stop training and
close the training unit have been detrimental to (our) linkages with NARS and have hindered
important gains in our capacity development role.

The importance given to the production of refereed Journal papers is too high in
comparison to the importance given to the impact produced by contributing directly to partners
through training.

However, against this, there is also the evidence that researchers have spent increasing time
on training in recent years (see Chapter 6.2) and that a high proportion of them consider
training an essential component of executing and refining their research. So the lack of
incentives described above may well have dissuaded researchers differentially. (See below
section 4.3 for more detailed analysis.)

The above paints a fairly consistent picture of the Centers' formal commitment to capacity
strengthening often expressed through training and learning activities which are, in turn,
perceived as relevant to the needs of the NARS. However as we have seen in practice Center
policies are not always consistent with the formal commitments expressed.

Capacity strengthening cannot only be addressed through training and learning. Capacity
also involves resources, equipment, management arrangements, policy support etc. Centers
endeavour to enhance the relevance of the training through various strategies to cover the
other capacity strengthening requirements e.g. by including them in collaborative project
planning. Evidence on this is given, for instance, by the increasing amounts of 'flow through'
funds managed by some Centers (e.g. up to 80% of a given project's funds goes to NARS at
ILRI). At the same time 72% of researchers regarded 'inadequate resources in NARS/NARI'
as constraining the take-up and impact of their research. Furthermore in the course of field-
work in SSA the Panel encountered many instances where those trained were unable to use
what they had learned because of lack of operational resources a reality affirmed by
questionnaire results, and discussed further below in relation to outcomes and impacts (see
Chap. 7).

A major concern expressed in the Center interviews concerns the relevance of present-day
training through project funding to longer term institutional capacity needs. In the short
term, project funding may help ensure that inputs such as equipment and operational
resources are provided to complement the training provided. But over the longer term, the
strength of the institutions may suffer because it has become more difficult to form a 'critical
mass' of researchers in a given area, or to form multidisciplinary teams who would sustain
research and be a force to influence institutional and political change. The importance of
these contributions is illustrated in some of the cases studies which had major impact at
institutional and field level (e.g. Bolivia, Case studies 1, 2). At the same time, projects are
frequently too short to accommodate higher degree training, which may be in the best
interest of the trainee and their institution.

4.3 Criteria for judging relevance

Relevance is generally judged by Centers in the context of their collaborative research
programs with NARS. Both CIP and IRRI for example regard training as relevant to those
areas of NARS research which are shared with the Center. Relevance here is both in relation
to implementing research (i.e. ensuring that data can be collected and field-trials organised)
and encouraging the adoption of new techniques and knowledge. This view is supported by
many researchers:

Training & capacity building are essential complementary to research and are
essential for enhancing food production and facing starvation in most of the
developing nations. Quality research cannot be implemented without qualified
staff members and therefore more funding and other resources need to be allocated
to these important activities. (Open ended comment in Researcher

A deeper analysis of questionnaire data as to the rationales of researchers for undertaking
training and learning activities, throws further light on their perceptions of relevance.
Researcher responses suggest that:
* Those who regard formal training as important for NARS capacity strengthening are
highly likely to regard skill shortages in NARS as a constraint on the take-up and impact

of their research (p < .000). On the other hand those who regard skill shortages as a
constraint may still spend less time than average on formal training.
Those who consider the lack of skills in NARS as an important constraint for research,
also consider informal training and learning as important (p <.002). However as with
formal training this does not mean that the researcher concerned spends a high
proportion of his or her time on informal training/learning activities.

These results tend to confirm that for researchers the justification for training and learning
activities with NARS is complementary to their research.

However, a much wider interpretation is assumed by some of the NARS. This is reflected in
the report of the recent internal review of ILRI's Capacity Strengthening Unit, which quotes
criticism from NARS representatives that the training is too project-driven, rather than
needs-based (Youdeowei et al., 2005):

'The majority of the training programs were not directly related to the
needs of the NARS programs. Rather, most of the training programs are
based entirely on ILRI's approved research projects. The effect of this bias
for ILRI's research program focus in training, has tended to limit the
impact of ILRI's CaSt activities on livestock development in the region.'

If relevance is assessed in terms of the extent of overlap of Center research goals and
associated training with NARS needs and the career needs of NARS researchers then
most CGIAR training can be judged as relevant.

However, when the needs of the NARS extend beyond Centers' research priorities, as they
often do, different conclusions can be drawn, as suggested by the ILRI case cited above.
Nevertheless, given Center mandates and funding, it would be unrealistic to expect a
response to the broader NARS' needs in such instances. Training outside the bounds of the
research agenda would, by definition, be outside the Centers' distinctive competence. But in
SSA, in particular, this poses strategic questions for the CGIAR as to whether more can be
done to reconcile poverty reduction (including the toll of HIV/AIDS) and the Centers'
mandates narrowly defined. The Panel heard different priorities voiced by NARS'
representatives as to criteria against which the relevance of CGIAR training should be
judged. Some clearly wanted Centers to respond to NARS needs even if they fell outside of
Center research mandates, often regarding Centers as among the few agencies with a
capability to respond to their needs.

Other problems of 'relevance' arise when the NARS re-orientate their priorities to match the
priorities of the CGIAR Centers. The concern here is: what does the NARS give up in order
to pursue priorities such as 'building capacity in molecular biology' in Africa? Is it to the
detriment of national institutions? Does it divert their efforts from what they ought to be
concentrating on? This can be regarded as a NARS problem of inadequate priority setting.
However when resources are very limited, quite modest funding can be sorely tempting for
NARS. Similar dilemmas face the weaker NARS in Latin America.

4.4 Priority setting in Centers and the NARS

NARS have different capacities, strengths and deficits, which also implies different capacity
strengthening needs. Furthermore capacity needs change over time as priorities shift and
countries and their NARS develop or experience setbacks. The ability of Centers to
differentiate between the needs of different NARS and to shape and adjust their inputs as
needs change, is therefore an important indicator of relevance.

The alignment of training and learning with NARS needs and priorities can be assessed in a
number of ways, including:
* The existence in Centers of training plans that are regularly updated and that specify
priorities at a sectoral and national/regional level;
* Regular consultation with NARIs and other partners as to priorities which may be both
formal and institutional or occur among scientists working together in networks or
collaborative research;
* Integrating training needs analysis into project planning.

Surveys of those responsible for training in Centers indicate that:
* Of the 12 Centers identified, 7 report that they have a training strategy or plan others
report that this is incorporated into broader Center strategies and plans;
* Most report that their strategies have been updated within the last 2 years;
* Regular consultation with NARS is rated as an important influence on these strategies.

However, those responsible for training (training Officers/focal points) are less confident that
consultation with NARS occurs in practice even if it is regarded as important: 17 out of 29
respondents said that 'regular needs analysis and priority setting with partners' did not
usually take place in their Center.

Since the demise of most Center training units and programs and the insertion of training
into projects, the processes in place to ensure the relevance of training have changed. Some
Centers that the Panel visited, such as IRRI and CIP, retain center-wide procedures for
assessing training needs and rationalising activities across subject areas and across regions.
Even in these cases, some difficulty has been experienced in applying these procedures
routinely, because of the decentralization of training. More commonly, needs assessment is
carried out at the project level and the effectiveness with which this is done is, consequently,
variable between projects within a given Center as well as between Centers.

Case studies at CGIAR Centers confirm these general findings. For example:
* There are well developed consultation procedures at least on paper -with NARS in
most Centers visited by the panel. Annual bilateral consultations; questionnaires to
NARIs; consultative groups or committees are said to be used to identify priorities.
* There is evidence that Centers shift the focus of their activities, between topics and
between NARS as needs change and as they respond to feedback. Thus IRRI has reduced
activities in Thailand and Vietnam but increased efforts in Cambodia and Laos.

The panel also noted in the course of field visits that well-documented procedures for
consultation and prioritisation in Centers are not always consistently followed, e.g. a
supposedly annual process might not be implemented for several years.

4.5 Factors shaping NARS priority-setting

Where Centers work with NARS to set priorities for capacity strengthening and training and
learning, they are dependent on the NARS' ability to undertake a national needs analysis
and set its own priorities. This does not happen effectively in all NARS. Thus in Bolivia, a
country that has experienced considerable political turmoil in recent years, the national
agricultural research institution (IBTA) was dissolved in 1998. Despite the creation of
decentralised, market-driven successor bodies, it was the view of country based informants
that there was now no 'voice' or coherent expression of demand across the country.
International research trends and project funding were seen as the main determinants of
training 'needs'. The lack of firmly articulated priorities also explains, at least partly, the few
cases encountered by the Panel where there was a perception on the part of the NARS that
Centers impose their priorities, or even make use of the NARS for carrying out their own
agenda (e.g. Ecuador Country Study). This kind of institutional weakness observed in certain
cases in SSA and LAC contrasts with the situation in the Greater Mekong Basin. In the latter
case, relatively strong NARS claim to have been able to articulate national priorities more
effectively. As one government official observed: 'Whatever training the CGIAR does in this
country is consistent with national priorities and has been agreed with (the NARI).'

It is difficult to generalise about the extent to which training priorities integrated with
research priorities become distorted by the availability of donor funding for projects. Survey
results for TOs and focal points, suggest that whilst donor priorities are not very important
(an aggregate score of 3.5 on a scale of 1-5) availability of funds is seen as more important
(4.3 on the scale). Case studies of Centers suggest that the non-availability of funds is the
most likely explanation of what occurs on the ground. From country based partners there
was more awareness of this actually or potentially occurring, and some evidence that it was
skewing the priorities of Centers. Certainly the highly erratic peaks in certain kinds of
training activity shown in Chapter 3 suggest a response to funding opportunities rather than
the result of systematic planning.

NARS ability to undertake needs analyses and put forward a coherent plan also interacts
with the security as well as the scale of funding. Strong NARS with secure own funding and
support at policy levels are better able to plan and prioritise than those without secure
funding or political support. Similarly those with longer term project funding from a donor
that 'is in it for the long-term' are better placed than those dependent on short-term funding.
Donors such as Rockefeller Foundation in Sub-Saharan Africa and the Swiss Development
Corporation in S.E. Asia and LAC were among those identified as supporting NARS over the
long-term and thus allowing NARS to develop reasonable planning horizons.

4.6 NARS' perception of relevance

Given the obstacles faced by both NARS and Centers in defining appropriate training plans,
the Panel collected evidence from various sources on the NARS' perception of the relevance
of the training actually carried out.

In the first place, as mentioned above, most of the NARS' trainees who responded to the
survey, undertook their training with encouragement from their employers, which suggests
confidence on the part of their institutions that the training would meet their needs. The
trainees themselves reported reasonably high rates of positive outcomes at the personal,
institutional and broader levels as shown in Chapter 7. However, as pointed out initially,
some positive bias must be included in these results, and the proportions of negative
perceptions varied from about 30% to about 60%, depending on the criterion and the region
These cases could be considered attributable to lack of relevance, but they were often
associated with a lack of opportunity to put trainees' newly acquired knowledge and skills to
use afterwards (7.2). So it is arguable that inadequate post-training provision and
inappropriate candidate selection were as much to blame as irrelevance of the training. A
similar interpretation seems valid for the different levels of trainee 'wastage' described in the
country reports. Thus, high levels of wastage in Ecuador or Malawi (DARS) contrast with
excellent retention rates of trainees in Bolivia (PROINPA) or Thailand for reasons more likely
to be related to institutional health than to different degrees of relevance of the training. This
is consistent with the 'model' outlined at the beginning of this chapter which suggest the
difficulty of isolating training and learning from the way provision is aligned with 'needs'
and ultimately the ability to use what has been learned. (This latter topic is further
elaborated in Chapter 7.)

Additional evidence was obtained in the Country Case Studies. Significantly, two cases
where the training was initially considered not relevant to local needs at all, were eventually
recognized to have highly successful institutional and field outcomes (Bolivia, Case 1;
Ecuador, Case 3). In others, there was some perception that the training satisfied the needs of
the Centers' research agenda more closely than the needs of the NARS (Ecuador Country
Report; Bolivia, Case 4). This contrasts with the very high degree of relevance evident in the
Kenya dairy case study (which became a model in tropical livestock production. While this
evidence is anecdotal, it raises again the issue of the precision with which NARS define their
needs, against which the relevance of the CGIAR training can be judged, but cases where
there is perceived to be a clear contradiction have, in the Panel's experience, been rare.

4.7 Conclusions

In general the Panel concluded that CGIAR Center training is broadly relevant to the
capacity needs of NARS. Centers are formally committed to capacity strengthening; and
many researchers within Centers as well as those with some responsibility for training and
learning (Training Officers and 'focal points') are evidently dedicated to helping NARS
strengthen their research base. It has also been argued by some researchers that relevance
has been reinforced in recent years by the decentralisation of training to Center researchers

who are now more closely involved in specifying training to match the needs of collaborative
research projects.

This broadly positive assessment needs to be qualified however, in three ways:
* First, what happens in practice does not always match formal commitments. There are
few incentives to become involved in training and learning according to a majority of
Center researchers. The close-down of training units and programs in many Centers has
reduced the ability of Centers to plan, coordinate and monitor relevance. Long
established processes of joint planning and consultation between NARS and Centers are
in some Centers now less used than they once were.
* Second, funding arrangements and in particular the growing dependence on project
funds can affect relevance. In some Centers project funding has been said to increase
relevance as researchers are now more committed to training and learning activities that
are integrated into collaborative research. However the short term nature of some project
funding can undermine NARS' capacity by reducing the time horizons for planning and
investing and by subsidising operational investments that are not sustained once the
project ends. Where NARS are weak and under-resourced it is also possible for Center
led project priorities to distort NARS own priorities pushing them in the direction
where funds are available.
* Third, judgments as to relevance depend on the criteria used. Judgments are most
positive if one stays within the parameters of Centers' research mandates. However
where the needs of NARS do not closely overlap with Centers judgments will be less
positive. This may be the case if NARS's priorities are broader than those of any one, or
all, the Centers, even though coinciding with broader CGIAR goals such as poverty
reduction or alleviating hunger.

The Panel takes the view that it is justifiable to assess the relevance of training within the
parameters of the Centers' research programs. This does not imply ignoring broader NARS'
capacity needs, but these must be addressed in collaboration with other agencies with a
different, complementary or more development-orientated mandate. The challenge is
greatest in SSA and a commitment to capacity strengthening in this region may require
innovative approaches to the delivery of training that goes beyond the strict requirements of
Center mandates. The most immediate way to improve relevance is put in place
standardised needs-assessment protocols across the full range of the Centers' collaborative
research projects. At the same time, there is an evident need to assist some NARS in the
establishment and articulation of valid priorities, which the CGIAR can then seek to
complement and support.


This chapter assesses the quality of teaching and learning in the CGIAR. It begins by
reviewing how quality is defined in the field of vocational education and training (VET)
considering both outcome and quality assurance (QA) approaches. This sets a framework for
the evaluation approach adopted in this study by the Panel. The chapter then considers the
QA methods that are used in Centers including how they are applied and to what types of
training and learning. Feedback from ex-trainees is discussed in terms of their satisfactions
with training quality and the utilisation of what is learned. Finally in the concluding section
recommendations are outlined as to how training quality might be improved.

5.1 Defining quality

Quality in education and vocational training is difficult to define, describe and measure for a
number of reasons. There are fundamental differences in approach between those who
favour an output model20 that looks for quality criteria against standards and a process
model that seeks to establish that procedures are in place to assure quality. Output models
confront questions about what standards and criteria to use knowledge acquired,
student/trainee satisfaction, competencies, usefulness in post-training settings; and whose
judgements count most: trainers, trainees, employers and at what point in time these
judgements are best made. (For example there are many evaluations in training institutions
including IARCs that depend on end of course responses rather than longer term follow-up.)
Process models follow training through from needs analysis to trainee selection, course
design including pedagogic aspects, delivery, feedback etc. These approaches also have
their difficulties although they have become the preferred approach to evaluate education
and training21. In particular linking processes with outcomes has in practice proved to be
difficult: how do we know whether a pedagogically 'good' course leads to better outcomes.

As was noted in the Inception Report for this study the diversity of training which includes
PhDs, Masters Degrees, training of trainers, group courses, experiential/informal and work-
based learning poses additional difficulties in assessing training/education quality in the
CGIAR. It was seen as unlikely that identical judgements could be reached for all the
different categories of CGIAR training and learning as has proved to be the case.

The 'model' adopted in this study is a pragmatic compromise among the alternative and
sometimes contested approaches referred to above. This consists of:
* An assessment of the ways in which Centers implement training and learning. This relies
mainly on an examination of the systems in place to assure quality from trainee selection
through to curriculum development, delivery and follow-up. This assessment draws
mostly on case-studies of CGIAR Centers, questionnaires to Training Officers/focal
points and where available EPMRs and 'impact assessments'.

20 The most well known example of an 'output' model is probably that associated with Kirkpatrick (1967)
although competency models that focus on the capability of trainees (Marrelli, 1998) have now become more
21 See for example the EU's 'Copenhagen' Process. (Copenhagen Process: First report of Technical Working Group
European Commission, Brussels, October 2003.)

* Feedback from trainees and partners. This includes feedback on their satisfactions and
assessments of quality as well as the reports on 'outcomes', e.g. usefulness of what was
learned for their subsequent work and careers. This relies mainly on questionnaires to ex-
trainees and partners and some contact with trainees in the course of fieldwork in
Centers and Countries.

5.2 Methods of quality assurance in Centers

The survey of Training Officers (or 'focal points' where no such role existed) were asked how
the quality of training was assured in their Center.

Table 5.1 Training officer survey: question 18

By what means does your Center assure the quality of the training it provides?
Feedback from individual learners 27
Feedback from partner organizations 19
Peer review of training materials 18
Feedback from University partners for PhD & MSc students 14
Updating trainers' methodological skills 11
Indicators as part of an evaluation system 9
Independent evaluations 7
Applying an explicit, written QA system 6
No explicit quality assurance 3

Obtaining feedback from learners, feedback from partners and peer review of training
materials were reported as the most common approaches to Quality Assurance (QA).
However there was much less priority given to obtaining feedback obtained from PhD and
MSc students than from course attendees.

In the course of visits to Centers the Panel was able to confirm that these means of QA did in
fact occur. However field visits to Centers and Countries also highlighted difficulties in
practice. For example:
* There was less feedback obtained from individuals and very little from those involved in
practical experience-based learning e.g. in field-stations or labs.
* There was virtually no feedback from in-country activities, which given decentralisation
and their importance in CGIAR training and learning, constitutes a major gap in
* Nearly all QA processes referred to, applied to course attendees there was little or no
QA for other forms of training, education or learning.

With the decentralization of training to researchers, the results of whatever feedback is
obtained remain in the scientists' domain and are not necessarily incorporated into
institutional measures to improve training quality.

The table above also suggests that independent evaluations play a minor role in quality
assurance. This was corroborated by the Panel. In their view, EPMR's have generally paid
very little attention to training quality or to processes in place at the Centers for monitoring

it. At the same time, Centers have made relatively little use of internally commissioned
reviews to cover this area, and those carried out have, in many cases, had two defects: first, a
lack of independence and second, the reliance on survey data without due recognition of the
positive bias in the results which this is likely to produce (see Annex III for a summary).

Those responsible for training and learning in Centers were also asked in questionnaires to
rate what they regarded 'as important to support training quality' and to contrast this with
what happened in practice. There were some notable discrepancies, particularly so for the
following items:
* 'Regular needs analysis/priority setting with partners';
* 'Training/learning expertise to advise on training methods';
* 'Training facilitated by specialists in adult learning';
* 'Screening of applicants to get the right trainees'.

All of these were seen as important for quality but not occurring in practice by a majority of
respondents. The absence of pedagogic expertise in training methods and adult education
is especially striking.

Deficits in quality were often attributed to the demise of training units and training officers
in many Centers. Strengthening training units is also seen as a priority by researchers
responding to the Researchers' Questionnaire Survey although from discussions with
researchers the kind of training units foreseen are different from those that previously
existed. Results for researchers who responded to questionnaire items on quality are shown
in the table below with their rating of factors seen as most important to raise quality ranked
from the highest to the lowest.

There is only limited agreement between researchers and those responsible for training as to
many of the ways that would ensure quality. Furthermore aggregate results for researchers
are less emphatically positive, probably because of differential levels of involvement in
training activities. Researchers are most keen on measures that involve them and less
enthusiastic for those that might imply an enhanced role for a training unit.

When visiting Centers the panel encountered many specific examples of good practice in
quality assurance. These were most evident in Centers that had retained some kind of central
Training Unit or function. Examples of good practice include:
* Involving a training unit/department in the design stage of research projects to clarify
learning objectives;
* Manuals and toolkits for trainers often geared to the needs and experience of
researchers who will be responsible for course delivery and made available also to
regional programs by HQ staff;
* Systematic feedback gathered from trainees at the end of courses and the maintenance of
accurate trainee records allowing for the periodic or occasional follow-up of alumni;
* Providing a resource person who is a pedagogic expert to facilitate researchers work
alongside them in preparing and delivering their courses;
* Setting up an electronic resource of courses and training materials, which can be
consulted, downloaded and re-used.

Table 5.2 Researcher survey: question 13.

Which of the following do you see important to ensure
training quality?*

Important Neutral Not Important
Opportunities for researchers to 48.9 24.9 26.2
update scientific content
Researchers involved in course 46.9 27.9 25.2
Screening of applicants to get the 46.5 23.5 29.9
right trainees
Standardized record keeping of 38.9 29.6 21.5
training and trainee related data
Systematic collection of feedback 36.6 25 38.4
from trainees
Regular training needs 33.8 31.9 34.3
analysis/priority setting with
Effective backstopping from 29.9 24.8 45.3
training office/unit
Training/learning expertise to 28.8 26.4 44.7
advise on methods and delivery
Development of best practice guides 27.5 24.6 46.9
for systematic use
Training facilitated by specialists in 26.0 23.0 50.0
adult learning
External evaluations of training 22.5 24.4 53.1
(additional to EPMRs)
* N=204-220 depending on item; values show % total replies for each line item

Examples of Center Good Practice

As noted above, these examples rely heavily on some kind of central training unit,
department or resource which nowadays only exist in seven of the Centers. In addition the
Panel identified major problems with the quality systems that are in place:
* Researchers are not required to follow guidance or advice and in some cases do not;
* Obtaining periodic feedback from subsets of trainees after course completion when
trainees return to work is not common, even though this is recognized as 'good practice'
and often more telling than feedback obtained on course completion;
* There appears to be little or no quality assurance systems in place for those involved with
degrees, on-the-job or informal learning even though these are major elements in the
CGIAR training and learning offer;
* There is a particular problem with 'quality on entry' of trainees due to deficient basic
training of applicants from many countries where the CGIAR is engaged (e.g. see
Country reports: Cameroon, Bolivia), as well as to lax selection procedures;
* Generally poor record systems for in-country trainees, with one or two notable
exceptions and very little follow-up at country level.

Many of these problems echo the findings of previous internal reviews carried out by CGIAR
Centers. (See Annex III, for Summary of Internal Reviews.) With regard to the quality of
training these reviews concluded:

The issue of how to implement QA (and quality control) for informal learning and training is
more challenging than for traditional training courses. However this is an undoubted
priority in the CGIAR and there are a variety of methods that could be adopted. Surveys of
partners and trainees used in this study are one method dependent of course on the
maintenance of contacts detail records. There are also instances of good practice already

emerging such as CIP's intention to incorporate learning objectives routinely into
collaborative research projects.

Comprehensive QA systems systematically applied are not the sole determinants of quality.
There are within the Centers visited enthusiastic researchers with their own innovative ideas
about learning who appear to inspire learners and adopt effective pedagogic methods.
However without effective systems it is difficult to consistently guarantee quality.

5.3 Feedback from ex-trainees, partners and NARS

Questionnaire surveys of ex-trainees and research partner were used by the Panel to obtain
feedback on the training and learning that has taken place. The positive bias which is likely
to occur is survey information is fully recognized, since the less satisfied would tend not to
reply. To counteract this, the Panel conducted interviews widely with alumni, partners and
their superiors in the countries which they visited, which gave a more representative sample,
albeit on a smaller scale.

A series of questions were asked in the surveys about trainee satisfaction both for those who
attended courses and for individual trainees. The overwhelming majority of trainees were
satisfied and many were strongly positive. However there are differences in levels of
satisfaction for different items. Course attendees for example were most satisfied with course
content, quality of teaching, organisation of course and quality of equipment, but least
satisfied with the balance of country specific and international content and the balance of
theoretical and practical knowledge imparted. The latter items raise particular quality

Table 5.3 Trainee survey: question 5

Satisfaction with Aspects of Course (Course participants*)
Areas of satisfaction Completely satisfied Satisfied
(%) (%)
Course content 55 91
Organisation of course 51 91
Quality of teaching 47 90
New training skills acquired 30 80
Opportunities to interact with trainers 36 75
Opportunities to interact with others on course 41 79
Balance of theoretical /practical knowledge 29 75
Quality of course material 44 87
Balance of international/country specifics 21 63
Quality of equipment 50 85
N is between 194/284 depending on item

Individual trainees also rated their training/educational experience positively even if
slightly less so than course participants.

Table 5.4 Trainee survey: question 6

Areas of satisfaction (Individuals*) Completely satisfied Satisfied
(%) (%)
Research opportunities at center 58 83
Support from supervisor or mentor 56 84
Cooperation with host university 37 63
Interaction with researchers at Center 38 69
Availability of equipment, facilities, resources 55 70
Learning/working with experienced researchers 37 69
Balance of international/country specifics 27 67
Availability/access to information/publications 56 82
*N approximately 170, depending on item

Fieldwork at Centers confirmed these aggregate results. For example in one focus group of
11 learners taking MSc or PhD courses, overall satisfaction was consistently high and
interaction with supervisors was especially praised. In another case, the trainees pointed out
that the content of their training was uniquely appropriate, because the IARC's are now
probably the only institutions worldwide where molecular genetics and traditional plant
breeding are dealt with in an integrated fashion. It is also significant that training quality was
very seldom brought up as an issue in the Country study interviews, leading to the
conclusion that it was generally considered to be satisfactory.

At the same time for individuals as for there are items which raise questions about aspects of
training quality. In particular there appear to be reservations about cooperation with host
universities, the balance of international and country specific content and the opportunities
to work with experienced researchers including researchers at Centers. The matter of
balance between international and country specific content, which featured for both
individuals and course participants, highlights the tension between the global role of the
CGIAR and trainee demand for 'local' or regional content. Ex-trainees in Vietnam and
Thailand touched on similar topics in the context of country-based training and the
likelihood that this would be more relevant to their needs than that delivered in Center HQs.

An analysis was conducted of the minority of trainees who were consistently negative (or
more precisely 'not positive') in their ratings of the training they had received.
Table 5.5 Trainee survey: analysis of negative replies

Difficulties using knowledge/skills x Positives/Negatives (Trainees)
Positives /Negatives Means N Standard Deviation
Negatives 3.69 68 .868
Positives 4.39 252 .730
Totals 4.24 320 .812
P < .000

This analysis showed that the single most powerful predictor of negative ratings by ex-
trainees was their difficulties in using knowledge and skills.22 These differences were statistically
significant across all outcome and quality filters. In brief:
* Dissatisfaction with training is greatest among those who report they have had too few
opportunities to use what they have learned.
* Negative ratings of training quality were also strongly correlated with few opportunities
to use what they had learned.

There is also a clear thematic or disciplinary divide in levels of satisfaction expressed by
trainees. As the table below indicates, the most positive ratings are made by those with a
background in Livestock, Fisheries, Crop Protection, Genetic Resources etc; and the least
positive among Social Sciences, Policy, Economics. Research Management etc.

The table shows a trend towards higher degrees of satisfaction in the biological than in the
social sciences. The differences were not always statistically significant, depending on the
numbers of observations, but in the larger classes of Crop Protection, Genetic Resources and
Crop Breeding, where the proportions of positives were 80% or more, these exceeded the
values for Economics, Policy or Social Science (67-71%) at levels of probability between 0.05
and 0.01. No differences were found involving NRM, another of the larger classes, or the
other classes with lower total numbers of responses.

Table 5.6 Responses by training theme

Proportions (%) of positive responses, by training theme23
Theme n Positives (%) Theme n Positives (%)
Livestock 30 87 NRM 104 77
Fisheries 9 86 Agroforestry 63 76
Crop Protection 125 85* Research Man. 61 75
Genetic Resources 144 85* Economics 49 71*
Forestry 28 82 Policy 35 69*
Crop Breeding 156 80* Social Sciences 43 67*
n= total number of responses
* Themes where proportions of positive responses are significantly i,. ..i. nt (P= 0.05-0.01)

These finding is open to several interpretations:
* Course content and training opportunities are better developed for the themes on the left
of the above table (e.g. livestock, crop protection, genetic resources etc.) than those in the
column on the left (e.g. social scientists, policy specialists, economists etc);
* The judgments of social scientist, policy analysts and economists are also influenced by
their need to become familiar with biological topics;
* Those who are negative come from disciplines more likely to be critical about courses;

22 Negatives were collated from different parts of the trainee questionnaire and respondents were scaled
according to the consistency of their dissatisfaction. This was then correlated with new multi-item variables for
'outcome' and 'quality'.
23 Note: Respondents were able to identify themselves with more than one theme

*There is also less of a correlation with opportunities to use what has been learned and
opportunities for using what they know among social scientists when compared with
those involved in crop-breeding. This may also have to do with the state of social
sciences in some NARS which offer limited research opportunities.

The partner questionnaire survey did not explicitly ask for satisfaction ratings or about the
details of training and learning quality, even though many were ex-trainees. However in the
course of country fieldwork which always involved interviews with NARI and NARS more
generally, consistently positive views were expressed by partners' representatives. Issues of
quality were not raised, but quality was assumed to be positive. This would be consistent
with trainee findings insofar as partners, by definition do have opportunities to apply what
they learn whether through courses, individually or informally.

A recurring theme in the Country studies was the value to local researchers of the informal
learning which occurred in the course of collaborative work, or due to the long-term contacts
established between local researchers and Center staff after formal training. Many perceived
this informal learning to be more important to them than the formal activities. Testimony of
this is given in many of the Case studies annexed to this report. (Annexes IX-XX; see Bolivia
Case Studies 2 and 4).

5.4 Conclusions

Ex-trainees were highly satisfied with different aspects of training quality, including course
content, quality of teaching, opportunities to interact with others etc. The minority of trainees
who were not satisfied appeared to be influenced by what happened after they completed
their training; not being able to apply what had been learned was a powerful predictor of
dissatisfaction. Given the highly subjective quality of these judgments the Panel would view
training quality to be generally but not uniformly good. More positive conclusions would
require confidence in a CGIAR-wide quality assurance system. Most of the views refer to
past training, i.e. before decentralization. Now researchers are more or less solely
responsible, and it is difficult to extrapolate from the past degree of satisfaction to the
present prevailing conditions.

QA systems for training even though they exist, are partial in their coverage and unevenly
applied across CGIAR Centers. Systems that are in place are not always implemented and
not all Centers have them. QA systems have been weakened by the reduction in specialist
training units or functions and the lack of pedagogic or adult education expertise among
Center staff. QA systems that do exist are applied mainly to courses. Informal training and
learning and individual training, both degree and non-degree, is not within their scope.
Country based training and in-country project based learning, more common because of
decentralisation and the integration of training into collaborative research, are not covered
by the QA systems that do exist. Although there are examples of good practice in Centers, it
is difficult to be confident that quality issues are being monitored and that systems are being
'steered' as a result. The Panel has concluded that at the very time that decentralised modes
of training delivery are challenging researchers to expand their pedagogic understandings,
there are fewer and fewer back-up resources available.

Explanations of these developments are often linked to lack of core funding. However they
can also be linked to a lack of prioritisation by Center management and by the CGIAR more
generally. More consistent and positive messages would have to be circulated within the
CGIAR for Centers to make training quality a priority area, in which they would be willing
to invest limited core funds and seek out additional project funds that could be used for
supporting training quality. Whatever the intention, Centers (and in particular those in
Centers with a strong commitment to training and learning and capacity strengthening) have
picked up messages from the days of the TAC onwards that what they do is not valued and
is seen as competitive with research priorities rather than complementary. (Even though
TAC's main argument was that training/learning was not the main bottleneck in NARS
capacity strengthening, which raised questions about the worthwhileness of expenditures.)
The belief that Center 'management' does not support and in recent years has reduced
support for training and learning is widespread. Such perceptions were reinforced following
ISNAR's closure and further reinforced in the course of recent discussions about proposals
from the Science Council on 'System Priorities'.

On the basis of examples of good practice identified and what happens routinely in some
Centers and for some target groups, it is possible to specify protocols for a QA system that
would conform to international good practice standards. Such a protocol would include
standards and norms for:
* Explicit training policies that set targets and link training and learning objectives to
research priorities;
* Procedures and criteria for the recruitment and selection of trainees agreed with NARS
* Course design including pedagogic guidelines;
* Pedagogic support and skills training for researchers in teaching and learning methods;
* Reinforcing the support/training of researchers by feedback from trainees at course end
routinely and for a sample at least at follow-up periods;
* The specification and monitoring of learning quality and effectiveness in informal
learning situations;
* The feedback of QA system results to Centers so that planning and prioritisation of
training and learning is improved.

All of the above ways of assuring quality would require the existence of training support
resources and expertise in Centers. This might not be equivalent to resuscitating an earlier
generation of 'Training Units'. Such resources would, for example, have to work in tandem
with researchers and in-country collaborative projects and be attuned more closely to the
priorities and needs of NARS actors. However given the continued high volume of training
and learning activity within the CGIAR it will be difficult without such a system to be
confident that this investment is being spent to good effect for enough of the time, in all
Centers and for all types of learning and training.


This chapter begins by discussing different understandings of efficiency, their applicability
in the case of CGIAR training and learning and how issues of efficiency have been
approached by the Panel. The chapter then considers resources allocated for training
purposes; coordination within and among Centers; evidence of economies of scale and of
specialisation. It concludes with an overall judgement as to current levels of efficiency and
what more can be done.

6.1 Understandings of efficiency

Definitions of efficiency at their most simple are about how money is used: the ratio of inputs
to outputs. More complex definitions elaborate more on the input or the output end of the
equation, without loosing touch with this basic formula. Thus the World Bank Independent
Evaluation Group refers to the 'extent to which objectives have been (or are expected to be)
achieved without using more resources than necessary'; and the Development Aid
Committee of the OECD defines efficiency in its evaluation glossary as: 'A measure of how
economically resources/inputs (funds, expertise, time etc.) are converted to results'.

There are a number of problems applying these definitions to training and learning in the
* First, there is little data available on 'inputs' in terms of budget, expertise, manpower,
courseware or classrooms. There is certainly no data that allows for a systematic
comparison of inputs over time and across Centers.
* Second, what constitutes training as has already been demonstrated is diverse and even
where data exists in aggregate terms for some periods of time in some Centers they do
not allow for the requisite degree of differentiation.
* Third, now that researchers rather than specialist trainers lead on most training and
collaborative research projects contain most of what constitutes informal learning, it has
become difficult to break down their time. Training and research activities are so closely
bound together as to be indistinguishable in terms of inputs and arguably purpose also.
* Fourth, what training and learning is attempting to achieve is similarly diverse and is
both difficult to isolate from particular settings (e.g. eco-regional locations; crops and
commodities; and techniques and know-how) and difficult to attribute in isolation from
the actions and inputs of many others.
* Fifth, the benefits to the Centers themselves of carrying out training are seldom fully
considered in discussions of efficiency.

This latter point is especially important. Efficiency cannot simply be assessed on the basis of
what Centers achieve to the benefit of the NARS. From the outset, the Panel believes it is
important that any discussion of efficiency should fully recognize the benefits of training to
the Centers themselves (as was touched on briefly in Chapter 2). In the course of their
Center visits, Panel members found convincing accounts of why scientists considered
training to be an essential activity for them, quite apart from the benefits to the trainees. It
extended their capacity to carry out research, improved the effectiveness of partnerships and
thereby increased research impact, kept them abreast of modern scientific developments and

in touch with reality at field level, and in certain cases, even facilitated access to donor

To partially overcome the difficulties identified, new data was gathered through
questionnaires and case studies, and existing data available statistics and reviews were
further analysed. However none of this allowed the Panel to undertake a classic input/output
efficiency study at a CGIAR 'system' level. It would have been possible to focus resources on
one or two specific cases but even this would in our judgement have had limited yield given
the integration of training and research and the many possible and actual outcomes of
training and learning.

Given the circumstances, the Panel therefore fell back, as elsewhere in this study, on a
relatively pragmatic approach to gauge efficiency. It concentrated on what Panel members
considered on the basis of wider experience were likely to be the correlates of efficiency,
* The way resources are deployed;
* Coordination and economies of scale;
* Concentrating on areas of specialisation or 'comparative advantage'.

It also sought the views of CGIAR stakeholders researchers, trainees and partners to clarify
how they understood efficiency.

6.2 Deployment and targeting of resources

As CGIAR Centers have undergone reductions in core funds and in particular in unrestricted
funds, they have reallocated their resources in response. There is a perception that training
has been a major target of cuts which has been associated with closure of some training units
or departments, closure of training programs, the integration of training into research and in
some cases the devolution of group training to national partners. Figures on funding of
training were difficult to obtain, but at the System level such data are available on research
"undertakings" up till 2002 (Annex V). These figures show a slightly increasing trend, but
there are known instances where they do not correspond with the information available at
the Centers and there appears to be no standard practice as to how staff time or indirect costs
are accounted for. In addition, it can be difficult to disentangle training costs undertaken as
part of research from the overall research budget. As pointed out in Chapter 2,
restricted/unrestricted funding data for training are not available from all Centers, so there is
no reliable basis at present for estimating either the System's overall financial investment in
this activity, or the real trends in'core' funds.

Cutbacks in dedicated training units have been reflected in reductions in the numbers of
persons with specialised training and adult education skills working in Centers. From the
responses of Training Officers to the questionnaire survey and to which the response rate
was very high it appears that such expertise is confined to only seven Centers. In CIAT, for
example, there were 22 professionals in the Scientific Training and Conferences Program in
the late 1980's, funded mainly from unrestricted core funds; 6 in the early 1990's and none
today with a specialised adult education background.

Despite the reduction in specialist pedagogic skills there has not been an overall reduction in
training activity (as was evident from the figures cited in Chapter 3). However the inputs in
training provision and in support for informal learning are now more likely to involve
researcher than specialist trainer time.

One piece of evidence on this score is that the proportion of researcher time devoted to
training has not apparently fallen over the last 5 years despite the various cutbacks reported.

Table 6.1 Researcher survey: question 4

Proportion of Researchers Time Spent on Training
Proportion of time spent on In the last 1-2 years 5 years ago (N = 175)
training (N = 275)
Less than 5% 16.7% 19.4%
5-15% 37.5% 43.4%
15-30% 31.3% 26.9%
30-50% 8.7% 5.7%
More than 50% 5.8% 4.6%

Researchers were asked to estimate the percentage of their time spent on different categories
of activity formal and informal training, research and 'other'.

Table 6.2 Researcher survey: question 7

Percentage Time of Researchers
Activity Mean % time Standard Deviation
Formal training 13.2 13.7
Informal training 11.8 9.9
Research 44.9 24.0
Other 26.2 22.3

The 'mean' responses among researchers revealed a high proportion of time under both
formal and informal categories, some 25% of time compared with 45% for research. It is
because of responses like this both in interviews and questionnaires that it is reasonable to
assume that most estimates of resources expended on training based on formal training are
underestimates. These figures also suggest that reports at system level of the CGIAR's
investments in training are understated. The likely reason for this is the consistent under-
reporting of informal learning and training activities, which are increasingly important in

Researcher questionnaire responses also clarify the perceived connections between formal
and informal training and learning.
* Those who consider formal training as important also spend above average time on
formal training (significant at the .048 level).
* Those who consider formal training important are highly likely to consider informal
learning as important and the converse is also true. (This is highly significant, p < .000).
* Those who judge the outcomes of training and learning as positive for capacity
strengthening see this as a combined effect of formal and informal means: it is only those

who consider both as important who regard training and learning as having a positive
outcomes for NARS (p < .01).

As pointed out above, detailed analysis of the deployment of resources to training is partly
difficult because of the way management accounts are kept. It is however possible to obtain
useful cost data in Centers something that we would recommend for future efforts to
monitor training efficiencies at a system level. Thus in one Center that has retained a central
training function we were able to establish that between 2000-2004:
* Training costs were split approximately 50/50 between research program and the central
training function;
* Over the same period core funds accounted for only 5% of the research program total;
* The central training budget was made up of a number of elements of which coordination
(including course administration and support) was only 10% the other elements being a
separately funded PhD program, generic short courses and ICT systems.

A number of efficiency question are raised by this example.
* How far are restricted or project funds able to be spent for purposes consistent with
research and training priorities?
* How far are core funds deployed to ensure that training activities are well-focussed?
* What are the costs and benefits of coordination?

In the particular example cited above it was consistently asserted by researchers and Center
management that donor priorities did not skew research priorities in approximately 80% of
cases and that donors were especially keen on training and skill enhancement. In another
Center however it was reported that there had been a change in the mix of training as a
result of funding reductions and re-structuring. There was now less disciplinary research
and training this being previously supported by core funds, and more of a move towards
commodity research and associated training.

Targeting and re-allocating resources is one indicator of efficiency. There are many examples
of this:
* Reductions in long courses and in courses in Headquarters;
* Increases in the number of short courses many in-region;
* Growth in informal learning integrated into research;
* Switching resources between countries depending on NARS needs;
* The growth of networks as vehicles for training and learning.

There are two reasons that resources might be re-allocated in these ways. First, Centers
might be responding to financial pressures. This would imply the primacy of the input side
of efficiency rather than outputs or results. A second rationale for re-allocating resources is
that NARS needs and contexts and the potentials for partnership have changed. In order to
achieve results different forms of training become salient. Examples of this would include:
* Increases in in-country training because of the identification of capacity needs in the
extension system and as a way of gaining access to more trainees at a lower cost per
* Shifts of resources between countries, following reassessments of their needs and
capabilities and consultation with NARS and partners.

Although the Panel was able to find many examples of this occurring the weaknesses
already identified in systematic joint needs analyses with NARS partners does not give
confidence that this always occurs.

On the other hand, the expected increase in efficiency due to some of these measures may
not be fully realized. For instance, while in-country training may increase coverage at lower
cost, there was ample evidence in the Case Studies of the distinctive value which trainees
attached to headquarters training, which extended well beyond the particular subject area in
question; thus, the values of headquarters and in-country training were not perceived to be
simply interchangeable. Second, field evidence certainly supported the growth of networks,
and they may have a specially critical role in combating the problem of high staff turnover
rates, as for example, due to disease in Africa. But at the same time, they can only prosper to
the extent that their individual members are strong and the Panel found evidence that the
weaker members may be at a special disadvantage (e.g. Ecuador Country Study).

Types of training and learning

For active researchers and leaders in technology transfer, there seems to be consensus that a
combination of training types fitted to their specific requirements will continue to be necessary. These
are likely to concentrate on specialized short courses, specialized non-degree individual training and
higher degrees. At the same time, evidence from Ecuador underlines the importance of informal
training and learning experiences, and of long term contacts with the centers. The advantages of the
networks should continue to be exploited fully, but their success depends on the stability of the
members and the extent to which they meet the needs of individual partners, particularly the weaker
ones, merits revision. A variety of training delivery modes will continue to be needed, with increasing
use of on-line materials and e-learning, but this must not be at the expense of deterioration in quality
in areas where practical experience is essential.

The proper selection of training and learning modes and methods of delivery is one
important determinant of efficiency. The Panel has not taken the view that short or long
courses are of their nature more efficient or that individual degree courses are better than
non-degrees or periods of work experience. Rather it has been assumed that different
training modes are suited for different purposes in different contexts. This is well
summarised in the Ecuador country report.

The key issue is whether systems are in place to choose between modes and to match
trainees to these modes. The evidence we have is that these systems do exist and examples
can be found that appear to work well. For example:
* One Center, that still provides training courses, has a clear anticipation of demand for
more short specialised courses and individual non-degree and degree training, but a
reduction in longer more general courses;
* Centers operating in SE Asia have adapted their 'offerings' to move from training to
collaborative networks;
* The selective use of e-learning and downloadable websites to support researchers
improve and systematise their training and support self-directed learning.

Again however the Panel is not confident that this can be said to be universal mainly
because as previously noted the skills available in pedagogics and adult learning are so
thinly spread across the CGIAR.

A closely related issue concerns the type of trainee to whom efforts should be directed. Here,
again, the Panel would guard against generalizations. Deficient laboratory technicians may
be a more important limiting factor than a shortage of well trained researchers, depending
on the circumstances. However, their field experience did lead the Panel to three tentative
conclusions. First, that interventions at the highest policy level are often an essential pre-
requisite to overall capacity strengthening and while these may not fall into the category of
formal training, the payoff could be extremely high. The CGIAR commands the status and
recognition necessary in many countries to perform such a role. Second, deficiencies in
university education have major implications for capacity development at all levels, not least
at the policy level and the level of candidate trainees. Major multiplier effects are foreseen
not only by supporting the universities' own training activities directly but also, very
importantly, by bringing them more actively into the research field (e.g. through
collaborative research projects). Thirdly, the Panel understands that farmer training may be
necessary in the course of developing methodologies, and also that, in the absence of
effective extension systems, Centers are drawn into this area as the only means of ensuring
that technologies reach the field. At present, given the state of Center data bases and the
potential perverse effects of indicator systems, there is a need to be cautious in interpreting
the apparent major increase in farmer training reported in Chapter 3. However a permanent
shift of resources in this direction would be a cause for concern, even if financed from non-
fungible additional resources, as it would not be unsustainable but may actually discourage
local institutions from assuming their responsibilities in the longer term.

6.3 Coordination and economies of scale

Coordination, both within Centers and between Centers is one predictor of efficiency. In
Centers without a central training function there is usually no coordination of training as an
activity and often no training strategy. In such cases it is difficult to speak of training
priorities or the benefits of coordination. Researchers often spoke of the reintroduction or
strengthening of training units as a means of increasing efficiency of training. However those
who responded to the Questionnaire Survey were ambiguous in their views about TUs.
Where TUs existed the majority of researchers (60%) wished them to be reinforced. However
where TUs did not exist only 39% favoured their reintroduction. Discussions with
researchers as part of Center fieldwork, suggest that they would be most supportive of
particular types of training units or functions, better adapted to a research-led training offer,
rather than some of the units now closed. It can be argued that Training Units have come to
be regarded as the symbol of a commitment to training activity by a Center. If that is a
reasonable interpretation then the main issue is the policy commitment of Centers to training
and learning and appropriate organisational arrangements to realise that commitment. This
implies the need to manage and coordinate rather than the re-introduction of 'training units'
per se. Such coordination will need to cover not only training activities within the Centers,
but also between other areas of capacity strengthening expertise (e.g. IT and
communications) available in particular Centers.

The Panel encountered many ways in which costs were spread over a higher volume of
activity in the course of field-visits to centers. For example:
* The collation in electronic form of training modules and materials to permit their re-use24
(e.g. IRRI Knowledge Base)
* Translation of resource material both electronic and hard copy into other languages
(e.g. CIAT Farmers handbooks)
* Disseminating methods, outputs and curricula developed in one region to other regions -
which is additionally efficient where it involves cost sharing with partners (e.g. CIP's
dissemination of disease diagnostics material from Bolivia to East Africa.)

These approaches to scaling-up, globalising and circulating knowledge and techniques as
widely as possible seem to be among the most consistently applied in the Centers visited as
part of this study.

Efficient resource deployment seems to depend to a great extent on the networking and
negotiation capacities of Centers to align donor and Center priorities; the coordination efforts
of those responsible for training; and the strategic use of core funds. At present this appears
to only occur in a minority of Centers in the CGIAR. Lack of coordination between Centers is
reported as a problem in the delivery of training by both by researchers and those
responsible for training (focal points and training officers). Coordination among Centers can
be seen as one way in which Centers might become more efficient and achieve greater
economies of scale and synergies. Researchers who answered questions on this topic
reported little evidence of coordination at many different levels including: disciplinary,
general and specific training themes or use of technologies to deliver training. Those
responsible for training also agreed that Centers could cooperate more in training materials,
training content and training delivery.

National and regional fieldwork undertaken by the Panel suggests considerable variation in
the extent to which coordination occurs. Thus in Malawi the lack of coordination was
specifically noted as it was in other parts of Africa.

Malawi: Integration needed for scaling-up

The move of each of the IARC's into working through dissemination and scaling up methodologies for
each of their crops and building the capacities to do so raises the problem that extension methodologies
are not being developed within a farming systems context. Crop specific extension and scaling up
methodologies make little sense once the work moves beyond the pilot stage. There would be value at
this stage of the work in Malawi for the IARC's to begin to integrate their work and the capacity
building and training initiatives thatflow from it

(Malawi Country Study)

24 This was an efficiency measure previously identified by ISNAR in: Anderson, J.R., et al Impact of ISNAR 1997-

On the other hand in Ecuador there appears to be more of a tradition of coordination.

Inter-center synergies in Ecuador

No evidence was found to -" ..I, ,t lack of coordination between centers in their training activities. In
fact, several examples were cited of how their efforts had been complementary. CIMMYT's on-farm
economic research, and associated training, in the 1980/90's, laid the foundation of what is now
considered to be the on-farm research culture in the country. This was later developed and
strengthened through CIAT's training and sustained collaboration in participative research, which is
now a recognized feature of INIAP's overall agenda (Case study 3) and has been further built up and
supported by CIP's collaborative work and training (e.g. in the FORTIPAPA project). A second
example concerns product processing and producer-consumer chains, pioneered through CIAT's
cassava processing research and associated training on the coast (Case study 2). It was strengthened
through workshops run by ISNAR, and further developed through the CIP-led market chain potato
network, PAPA ANDINA which has strong t,0ninig I, ,1 i ii j components. The producer-market-
consumer chain concept is now well incorporated into INIAP's research policy for all crops. A third
example relates to the collection, description, conservation and exploitation of native plant and forest
species within INIAP, which has been supported through training and collaborative projects by
IPGRI, CIP and CIAT. One feature of all these examples is that the Centers' policies and approaches
to research and development are perceived to have been consistent and mutually supportive.

(Ecuador Country Study)

6.4 Specialisation and comparative advantage

One suggested measure of efficiency is the extent to which Centers confine their training
activities to those topics where they enjoy a comparative advantage.

In national fieldwork there was a consistent understanding of what CGIAR Centers had to
* Integrated approach to solving problems of world importance (hunger, poverty, resource
conservation), integrated across biological and sociological disciplines, and across
upstream' and 'downstream' levels of science;
* Long-term experience in the production and utilization of the mandate crops in the social
and physical environments where they are grown;
* Unique collections of germplasm and related institutional knowledge;
* Worldwide network of collaborators;
* Capacity to act as apolitical 'honest brokers' and facilitators internationally and inter-
* Excellent research infrastructure, documentation and information facilities.

Those responding to the training officer/focal point questionnaire were also clear about their
Center's comparative advantage. Thus the 'link to strategic research' and 'scientific and
practical experience in mandated area' was highlighted. However it was acknowledged that
in some instances training outside a Center's area of comparative advantage does take place.

For example:
* Occasional seminars on how to develop project proposals, given to network members as
a way to strengthen the networks;
* Students/scholars from IT area get trained by IT experts, while working on topics 'core'
for the Center;
* English course so that researchers can participate in the international scientific
* Experimental design, data collection, management, analysis;
* Scientific writing and Presentation skills.

Various explanations for these activities have been put forward:
* The absence of alternative suppliers say in a particular region or country;
* Such training is integrated with other training as a relatively small element and it would
be disruptive and expensive to insert another supplier for a particular module;
* It opens up useful networks for wider Center activities (e.g. research, dissemination, etc.);
* Difficulty obtaining English tutors who are familiar with the language of agricultural
science, making it desirable that Centers at least 'source' language tutors even if they do
not deploy their own scientists.

There appears to be an awareness of this issue and the related 'international public goods'
criteria among those interviewed in this study. Some see greater cooperation among Centers
as a way forward:

'Training activities on agricultural policies and marketing though relevant is hard to
approach from agroforestry standpoint. A coordinated CGIAR approach is better.'

On the other hand some informants wished to emphasise the positive aspects of these non-
core types of training:
* An unavoidable aspect of training where remedial elements often have to be added to
core curricula to even out gaps in trainee knowledge.
* The importance and benefit to CGIAR Centers of improved partner ability to raise
funds in a specialist funding market (hence fundraising).
* The equal importance of English as a language of scientific communication in an
international scientific community.

One of the training I have received was on scientific writing, including proposals. This
Course gave me the opportunity to be more realistic in research. I can now exploit different aspects of
my work to enhance my institution's image through publications, for instance. Furthermore, I can
now prepare scientific proposals even if I have yet to learn in that field to be more efficient.

Trainee Questionnaire, open ended responses

It would in any case appear from the data in Chapter 3 that the volume of such training
activity is a very small part of the overall portfolio of training and learning that is on offer
from CGIAR Centers. Thus, the whole Methods category accounted for 10.7% of group
trainee days, and about one third of this was devoted to statistics and experimental design,

areas in which experts in the subject matter (e.g. crop or animal scientists) are recognized to
be more effective teachers than experts in statistics. (Table 3.5 and Annex VII). Also, most of
the English teaching has been carried out by a single Center (Annex VIII). The
corresponding figure for individual trainees was far lower (3.2% total trainee days devoted
to Methods, Table 3.6), indicating that they were exposed to a very minor degree to possibly
non-core' subjects. A very similar picture is given at a country level, taking Ecuador as an
example, where it was estimated that at the most 4% of all training offered by CIMMYT and
CIAT was in areas not covered directly by their mandates and that might be considered
better delivered by other providers (see Country report, Table 2).

On the other hand there appears to be an extent to which Centers are driven to compensate
for inadequate trainee preparation by remedial inputs outside the scope of their mandate.
Field work in several of the countries visited drew specific attention to the deficiencies of
basic and university education, and to the effect which this had on the initial levels of
preparation of training applicants. Accordingly, for 67% of respondents to the Researchers
Questionnaire 'Selecting trainees more carefully' is seen as an important way to improve
efficiency and effectiveness of a Center's training and learning.

This is elaborated by researchers in open ended questions:

The level of the trainees is too low, they need to get a higher degree or go to a better school
first. It is not (Center X) job to provide general training on statistics, data entry

There is a lack of control over selection (quality) of persons trained [which] can create to a
large supervision burden with little return.

In some training activities there is a tendency to incorporate students that do not fit within
course requirements. Some are there for political or institutional reasons.

These problems were encountered in both Latin America and Sub Saharan Africa. The Panel
would also draw attention here to another aspect of candidate selection, which affects
efficiency. Evidence in Chapter 5 showed that NARS' satisfaction with training is strongly
related to how far it has been put to use afterwards and, as shown in Chapter 7, lack of post-
training resources has been a widespread limitation. The problem is recognized by the
Centers, but some of those visited were not comfortable with the prospect of assuming a
stronger role in imposing criteria for candidate acceptance. The situation has improved
somewhat with the insertion of training into research projects and since NARS have more
commonly had to pay for the training received. Nevertheless, the Panel believes that this
issue should be discussed frankly between Centers and NARS, and that the latter would
welcome norms designed to safeguard their own investment in training by ensuring
adequate post-training opportunities for their candidates. Such discussions might also form
part of more general discussions with other donors who might be encouraged thereby to
align their funding initiatives for capacity strengthening to NARI and NARS with CGIAR
training and research plans. This would enhance the efficiency of the overall training
process from the Centers' point of view, and perhaps also help reduce the levels of trainee
'wastage' described in this report.

Problems of trainee quality highlight broader problems of NARS capacity. These include the
state of Universities, government's policy commitment to agriculture, the funding available
to NARIs for operational costs and the limited ability of Centers or indeed the CGIAR as a
whole to address this scale of problems. Although many of these problems will have to be
addressed by others e.g. donors, governments and universities, this also highlights the limits
to what individual Centers can achieve on their own. This harks back to questions of inter-
Center coordination, discussed earlier.

Discussions of where the Centers' comparative advantage for training lies, raises issues
about whether other institutions are deemed to have comparable or superior capacity in
what have hitherto been regarded as Centers' own 'core' areas. Clearly, the Centers'
advantage changes as their research evolves, and other suppliers acquire new strengths. In
this context, the 'devolution' of training activities to other suppliers, including the stronger
NARS, is frequently called for. The Panel did not come across examples where this seems to
have worked successfully. Rather, they were impressed by the case at Egerton College,
Kenya, where despite extraordinary preparations by CIMMYT, collaborating donors and the
College itself, the numbers of candidates for the production course have dwindled badly in
recent years, not for lack of demand but for lack of funding. The causes are probably
complex, as is the general issue of how far the stronger developing country institutions can
successfully take on the support of weaker neighbours, or would be welcomed for doing so.
In any case, the main onus would seem to be on the NARS themselves to ensure that their
trainees are sent to the institutions most suited to their needs. Previous distortions which
arose when training costs were covered completely by the Centers should now be largely
removed. The most promising future strategy for efficient sharing of responsibilities would
seem to be through the multipartite training partnerships, already in operation, where
northern and southern institutions are linked with the Centers, and the work load shared
efficiently according to the distinctive competence of each one.

6.5 Conclusions

The pre-requisites for the efficient management and delivery of training and learning are not
in place in most Centers. It is therefore difficult to assess overall efficiency. There are many
examples of 'good practice' but these are unevenly distributed. The most important deficits
are inadequate pedagogic and coordination resources within most Centers and the absence
of systematic financial and monitoring data. However it should be emphasised that the true
efficiency of training and learning is its contribution to the effectiveness and take-up of
research rather than considering training in isolation.

Investment by the CGIAR in training and learning through formal and informal means
continues to be high. Most training of whatever type is delivered by researchers many of
whom although enthusiastic teachers, have limited pedagogic experience, whilst skills in
teaching and learning, curriculum development and trainee follow-up have become scarce in
most CGIAR Centers. Given the close integration of training and research it is inevitable that
training and learning will continue to be an important and resource intensive activity in the
CGIAR. In the past TUs have also contributed to planning and coordinating Center wide
training activities, as well as to the retrieval and adaptation for widespread dissemination of

training materials. At present this does not happen consistently or widely enough in most
Centers. The Panel takes the view that given the scale of resources deployed there is a need
for more consistent coordination, backstopping, advice and support in all Centers.

The lack of coordination between Centers is also a problem especially in Africa. Synergies
could be achieved if there was more inter-Center cooperation but this would also depend
on policies and resources within Centers (or in decentralized country programs) to be able to
manage this effectively.

As previously noted (see Chapter 5 conclusions) the Panel does not favour the reintroduction
of traditional TUs, it does take the view that both coordination and pedagogic support are
needed in all Centers. This could be organised in various ways and will need to reflect the
specific mandates of Centers and their decentralised in-country activities.

In some areas Centers have evidently adopted efficient practices. This would apply to the
way training 'products' are usually turned into generalizable, 'global' goods thus achieving
economies of scale in their production and use. There is also evidence that in response to
changing funding levels and NARS needs and priorities, Centers have re-allocated resources
between types of training, countries and themes. However given the unevenness in joint
planning and needs analysis with NARS the Panel is not confident that these reallocations
are always planned in the most strategic fashion. There is also room for clear exchanges with
the NARS on the issue of candidate selection and likely subsequent deployment, to ensure
that they not only come with suitable pre-training preparation was also with adequate
possibilities of putting their training to use afterwards.

In general the Panel is confident that the overwhelming part of training and learning is
covered by Center mandates 'they do what they do best'. Exceptions can usually be
justified in terms of particular circumstances. However there is a proportion of non-research
related training activity, for which this is not so, for example where Centers try to cover
resource shortages in NARS out of project funds that cannot be sustained or where trainees
without adequate preparation are selected. These instances point to more generic capacity
issues than Centers and their training programs can address single handed and raise
questions not only of coordination among Centers but also of coordination with other
stakeholders, especially governments, donors and universities.


This chapter begins by clarifying the way the Panel defined effectiveness and linked notions
of outcome and impact. It then discusses aggregate responses from questionnaire data and
important regional differences in context or 'scenario'. Country studies are then discussed
allowing for a more detailed consideration of key issues including: CGIAR investment in
capacity strengthening in NARI; continuity in CGIAR interventions within the 'project'
mode of funding; the apparent preconditions for success; and the sustainability of outcomes
and impacts. This is followed by overall conclusions.

7.1 Understanding 'effectiveness', outcomes and impacts

As in other parts of this study, the Panel faced choices of definition with regards to
effectiveness and the related concepts of outputs, outcomes and impacts. In general the Panel
has followed conventional definitions. Effectiveness is usually defined in terms of the
achievement of objectives; and outputs, outcomes and impacts are intended to capture the
shorter, medium and longer term aspects of results. However the nature of this domain still
leaves open scope for different or specific interpretations. In particular the Panel considered:
* The parameters of 'effectiveness' in capacity strengthening;
* Criteria for judging outcomes and impacts;
* What sustainability means;and,
* Effects of the wider context.

The parameters of 'effectiveness' in capacity strengthening
Capacity resulting from training and learning is frequently understood within CGIAR as
individual killing or education, largely within a human capital framework. The Panel
starting from a capacity strengthening standpoint has adopted a broader view. Thus it has
examined how far acquired skills and capacities are actually used as well as acquired. It has
also looked beyond individual advancement, focusing where possible on organisational
benefits and the benefits of networks and inter-organisational linkage and how far these
have been sustained.

Effectiveness in terms of capacity strengthening has been regarded as a 3 stage process:

Figure 7.1 Effectiveness of capacity strengthening

Acquisition of Deployment of Sustaining skills
skills and skills & capacities & capacities

The feedback loop in the above figure is important because it highlights the consequences of
not sustaining skills and capacities once acquired. Instead of building on previous training
investments, a Center can find itself simply replacing and gap filling the basic skill set of a
new generation of scientists. This is explored in this chapter in terms of different regional
contexts or 'scenarios'. It is only possible to judge the effectiveness of Center training by

recognizing that contexts differ and shape what is possible to expect from apparently similar

Criteria for judging outcomes and impacts: synergies and trade-offs
Capacity is intended for a purpose. Both Centers and NARS expect that enhanced capacity
will encourage research that improves agricultural performance and raises the income of
farmers, whilst usually increasing national income and competitiveness and often reducing
risks of environmental depredation. Outcomes and 'impacts' have therefore been assessed at
several levels: individual, institutional and in terms of wider agricultural and socio-economic
goals. The Panel had has similarly tried to keep in mind two sets of (presumably) linked
criteria: the benefits to NARS, farmers and Countries and the benefits for Center research, its
take-up and further development. Even if not all training and learning will fully and equally
exemplify both sets of criteria, the synergies and trade-offs of each have been kept in mind.

Defining 'sustainability'
Outcomes and impacts highlight the dimension of sustainability or duration. Especially in a
project funded setting 'success' can easily be treated as a snapshot at the end of a project
cycle, irrespective of what happens subsequently. As has already been noted one of the
possible downsides of funding training and other capacity strengthening actions out of
project funds is that outcomes and results are not sustainable. However the meaning of
sustainability is not always straightforward. It can be interpreted as continuity of what has
been achieved, but it can also be interpreted as a more diverse set of outcomes left behind by
a particular project in which training or learning measures were an important part. The Panel
has chosen to take this more diversified interpretation of sustainability including follow-on
and spin-off outcomes as well as end of project results. The findings of case-studies in
particular support this stance.

Effects of the wider 'context'
Fieldwork and data-gathering in the NARES emphasises the reality that the CGIAR is always
operating within a wider context. For example a Center is only one actor among many in
developing countries. Objectives, purposes and intentions related to NARES capacity are
shared among many stakeholders and little can be achieved without stakeholders working
together. Outcomes and impacts are therefore not the result of what Centers do alone.
Although the Panel has selected cases and countries where the CGIAR has invested heavily
in training and learning or where training and learning appeared to be critical inputs,
fieldwork has demonstrated that it is often wiser to speak of the 'contribution' that Centers
have made rather than to seek to attribute results to Centers alone. This is especially so in
much NRM/systems research, where there are many collaborators and the issue is not so
much outcome/impact at the personal or institutional level, as what they have achieved
between all of those involved. Second the wider context focuses attention on a wider set of
contextual factors that make 'success' more likely. These include not only stakeholders but
also previous investments, government policies, donor priorities, local leadership, university
quality, international competition and public sector reform to name just a few.

7.2 Regional 'scenarios' and aggregate responses

In general the three regions within which the Panel has concentrated its NARS-specific
efforts appear to exemplify quite different contexts or 'scenarios'. These might be
characterized as:
* Unstable: Some countries in Latin America and SSA have been subject to considerable
political, and institutional, instability which has affected the deployment and
sustainability of CGIAR investments in training as well as the possibility of establishing
and sustaining partnerships. Even when individual skills exist, NARS capacity is
unevenly distributed and fragile.
* Under-resourced: Sub Saharan Africa has been affected by poverty, structural-adjustment
policies, limiting public investment, limited private sector resources and the
consequences of Malaria and HIV/AIDS. There is some similarity here with conditions in
parts of LAC represented by Bolivia in this study. CGIAR Centers have often found
themselves replacing previous training investments and existing skills have often been
under-utilised. In some countries NARES are often too under-resourced to define their
own priorities or support partnerships.
* Rapidly developing: In Asia and in this study notably in the Greater Mekong Basin sub-
region there has been and continues to be rapid development in the agricultural
sector and in the application of agricultural research. NARS capacity has strengthened
and national institutes, universities and the private sector have taken over many research
and training functions previously the province of the CGIAR. Partnerships with CGIAR
Centers are strong and research agendas are self determined.

These scenarios are not completely self-contained. Structural adjustment policies have
affected Latin America as well as Africa and there is considerable overlap between the
conditions in poorer LAC and SSA countries. Nonetheless it is also true that in LAC there
were within living memory stronger NARS than now exist, whereas in parts of SSA this has
not been so. The consequences of these 'scenarios' are easily masked by aggregate survey
results, but become clearer in more detailed case-studies. However there are a number of
aggregate indicators of these different scenarios. These include:
* Institutional stability: in Bolivia there is no national agricultural research institute since
IBTA was abolished in 1998. In Ecuador the Ministry of Agriculture and Livestock's
extension service was closed some 15 years ago and INIAP has faced funding and
institutional problems (see Ecuador and Bolivia Country Overviews). Malawi with acute
resource constraints has also faced great institutional difficulties.
* Agricultural GDP R&D levels are often low: Ecuador 0.26%; Malawi 0.75%; Thailand
* Labour turnover, the potential 'wastage' of skills varies across NARS and NARI, as the
following suggest: Kasasart University moderate25; Thai Department of Agriculture -

25 Figures, available in case study reports, are based on different sub-sets of staff, different periods of time and
different ways of measuring turnover. These are therefore estimates backed up by reasonably sound indicative

low; Bolivia PROINPA low; Ecuador INIAP high; Cuu Long Rice Research (Vietnam)

At an aggregate level respondents to the ex-trainee Questionnaire Survey are
overwhelmingly positive about the benefits that have followed from CGIAR training. Many
positive outcomes at a personal level and institutional level, though not all, were rated as
'important' or 'very important' results of training received. The same was true of questions
pertaining to wider CGIAR Goals, which were all rated as important or very important in
enabling respondents to contribute to broader CGIAR objectives.

Table 7.1 Trainee survey: questions 8, 9 and 10

Results in terms of personal, institutional and broader goals
Important/Very Important
Which of the following resulted from your training at personal level?
Taking on new tasks with higher responsibility 64
Increased ability in research priority setting and problem orientation 73
More research output (innovations, publications) from your work 58
Increased participation in collaborative research activities 63
Encouraged to undertake further training and education 43
Increased skills in project planning & fund raising 49
How important was your training in enabling you to contribute to
changes at the level of your institution?
Incorporation into research networks 45
Improved priority setting 56
New inter institutional linkages 49
Better access to information 55
Funding new projects 41
Better access to information 56
How important was your CGIAR learning experience to enable you
to contribute to the following broader objectives
New scientific knowledge 75
New attitudes and technologies 78
Farmers/consumers benefited 62

It is noteworthy, however, that trainees did not see as an important result of training 'finding
a new job outside of your country' although such trainees may well not have responded to
this survey. It can be argued that respondents to questionnaires were likely to be positively
disposed towards the training and learning they had experienced, however other sources of
information (e.g. NARS interviews) confirm these responses as representative of those
CGIAR alumni who have remained within their national agricultural research system. For
many ex-trainees the outcomes of training are seen as positive, key events in their
professional lives and a door opening to an international research career.

Nonetheless there are important regional differences which further illustrate the different
scenarios encountered. Trainee respondents were divided into those who were consistently
(over many items) positive in their responses when rating outcomes and those who were
either negative or at least not positive. This was done for assessments of personal outcomes,

institutional outcomes and in terms of outcomes about contributing to broader CGIAR goals
(new scientific knowledge, farmers benefited etc). The results are summarised below.
Table 7.2 Trainee perceptions of training outcomes

Personal Benefit (%)*
Negatives Positives

Institutional Capacity (%)
Negatives Positives

Wider Goals (%)
Negatives Positives

APO (N = 78) 37 63 32 68 35 65

LAC (N = 159) 39 61 48 52 31 69

SSA (N = 108) 60 39 56 44 49 51

Other (N =10) 30 70 30 70 40 60

Significance .002 level .007 level .023
Chi Square
*All percentages are of regional responses

The results show a hierarchy of judged effectiveness with APO coming ahead of LAC which
is in turn ahead of SSA in both personal and institutional benefits, even in terms of
contribution to wider goals SSA lags behind the other two regions. These results are
statistically significant.

'Tracking-studies' were undertaken at institutional level and country levels as part of NARS
fieldwork. One stream of activity was directed at NARI in order to establish what input
CGIAR trainees had made both at leadership levels and more generally. In terms of
In Ecuador INIAP, the national institution responsible for agricultural research and
extension has had approximately 400 training 'inputs' from CGIAR Centers. INIAP's
Director General, 6 of the institute's 10 Directors, 17 of 28 Heads of Program, and 13
Heads of Department/Units are CGIAR trainees.
In Bolivia PROINPA26, the General Manager and 7 of the 9 Heads of
Units/Regions/Scientific Programs are CGIAR alumni.
In Vietnam, VASI (Vietnam Agricultural Science Institute) 48 out of the complement of
approximately 480 scientists have been trained by CGIAR. Of these 18 are in senior
positions including the Deputy Director General, the Vice Directors for Plant Genetic
Resources, Root Crops and Hybrid Rice, Acting Director Legume R&D and the Head of
In Thailand, Department of Agriculture, (Ministry of Agriculture) 48 staff were trained
by the CGIAR. Of those with degrees or postgraduate degrees (33 individuals) 24 remain
with the department, many as Heads of Departments or Deputy Directors of research

26 PROINPA Foundation is a Bolivian non-profit organization oriented to promote technology innovation and the
conservation, use, management, and development of genetic resources

Similar findings as to the leadership roles of Center trainees have been cited in other studies
of training outcomes27.

National context is extremely important for sustainability however, as was suggested earlier.
In Thailand a relatively stable and rapidly developing country with a well developed NARS,
249 out of 541 names from a list of CGIAR ex-trainees (1995-2000) were tracked. Of these:
* 148 were found to be still working in the same field in the same organisation;
* 40 were found to have retired or died whilst in the same organisation.

Thus 75% of those tracked had remained in employment in the organisation where they were
based when trained.

The picture is not uniformly positive however:
* In Bolivia which has undergone political and institutional instability over the last decade,
many of those trained by Centers are no longer working in agricultural research for
example 12 of the 18 scientists trained by CIAT in one research station are no longer in
post and over a third of those trained in participatory research in Bolivia are reported to
be no longer utilising their skills.
* In many SSA countries (exemplified most obviously in this study's NARS by Malawi) the
combined effects of poverty and structural adjustment policies have constrained the
ability of governments to invest in agricultural research. Together with the effects of
HIV/AIDS this has undermined both the deployment of acquired skills and capacities
and their sustainability.

The aggregate results are somewhere between these positive and negative examples. Among
ex-trainees responding to the Questionnaire Survey, 55.7% reported that they continued to
work in the same organisation as before. This result would obviously be biased upwards
because fewer of those who had left would have been contacted.

It was previously noted that dissatisfaction with training is greatest among those ex-trainees
who report they have had too few opportunities to use the knowledge and skills what
they have acquired. This was probed further in the Trainees Questionnaire Survey. The most
prominent explanations for the non use of skills were resource related. Of those responding:
* 19.5% referred to lack of operational resources;
* 21.9% referred to lack of resources to support networking with relevant scientific
* 19.6% referred to lack of facilities and equipment (e.g., computers, lab facilities).

Further analysis indicates a regional effect here also: resource problems are most likely in
SSA and (to a lesser extent) LAC than in Asia. Statistically there are significant differences
between regions in terms of using what has been learned with 'no problems'. This is more
likely to be the case in Asia and Latin America than in SSA.

27 See for example Richmond et al.1998, In depth review of IPGRI's Documentation and Information on Training
Activities; and Raab et al.1999, The Impact of IRRI's Training Program: A different perspective.

7.3 Partners, training and 'results'

A particular and important result of training is the ongoing professional links that are
established between NARS scientists and CGIAR centers. Ex-trainees were asked about the
kind of contacts maintained with Centers where they had obtained training and education.
Table 7.3 Type of contacts with CGIAR centers and scientists

Ongoing contacts maintained with Center (N=251)
Maintained ongoing professional links with one or more Center scientist 209 (52.1%)
Undertaken collaborative research with the Center 131 (32.7%)
Undertaken a further course with this or another international Center 61 (15.2%)

The overlap between 'trainees' and 'partners' is evident when talking to CGIAR Center-
based researchers and senior managers. For some indeed the purpose of training and
education is to recruit partners whilst for many researchers trainees are recruited from the
ranks of those who are already partners. This is borne out by responses to the Partners
Survey, which confirms the high proportion of partners who undertake training or obtain
degrees in the course of collaboration with the CGIAR. An equally striking feature of the
partners' responses is the importance they attribute to informal training within a

It is reasonable to assume that the growth in informal training and learning within Centers is
associated with the growth in partnerships and networks often themselves involving
upscaling, 'adaptive' research and multiplier effects that engage with extension and
education systems as well as with policy actors. The importance attributed to informal
training is also consistent with other evaluations undertaken for centers. For example an
impact study of ILRI's graduate fellows program (previously cited) reported that 'working
with others' was considered by far the most important source of scientific knowledge for

Partners capable of participating in collaborative research are one of the most important
'legacies' of CGIAR training and learning activities. The goodwill towards Centers that result
from this relationship is striking in countries visited. Ex-trainees especially because of the
senior positions they often occupy are willing to open up research opportunities, insert
Center priorities into their own professional circles and perform a host of collegiate roles -
from meeting visiting scientists at airports to being positively disposed to joint funding
applications. A note of caution, however, is in order. The reduction in degree training in
some countries (e.g. Vietnam and Thailand) and associated funding opportunities since 1995
means that in some countries CGIAR alumni are ageing and often approaching retirement.
It was widely recognized that the CGIAR no longer offers 'free' training nor subsidies or
grants to the NARS to anything like the extent it once did. One partner even anticipated
having to pay for germplasm in the future. As a result the CGIAR may no longer be looked
upon as frequently as the obvious partners for collaborative research, and some evidence
was found in the field studies that in some instances Centers are now perceived more as
competitors for funding than as partners. This may have consequences for the future
resources available to the CGIAR for networking and collaborative research, and ultimately

for its capacity to leverage large scale research, from what is often a relatively modest
research budget.

Partner organizations include universities, NARI, regional and sub-regional bodies,
NGOs/CSOs, agricultural extension and farmers' organizations. Capacity strengthening can
therefore take many different forms. Partner respondents were asked to identify the main
changes that resulted in their organisation from training and education.
Table 7.4 Partner survey: question 14

Main changes for partners' organizations (N=148) percentage (%) responses
no change some change great change
New organisational skills and competencies 10.4 54.5 35.1
have been acquired
New priorities have been formulated 20.6 48.1 31.3
Organization's resources are now allocated 42.9 48.7 8.4
Enhanced role in networks 11.7 52.3 35.9
Enhanced inter-institutional linkages 6.1 52.3 41.7

It is noteworthy that although changes at an intermediate level ('some change') are reported
in all categories, the strongest changes appear to be in relation to networks enhanced inter-
institutional linkages and the weakest in the extent to which there have been changes in
how partner organizations' resources are allocated.

Partners were also asked to identify 'the main changes for the take-up and outcomes of
research'. Here also changes were reported under all the main categories offered:
* New research networks have been established;
* Knowledge and techniques are now more widely available;
* Knowledge that was not previously applied is now being applied;
* Knowledge has been adapted to specific settings, farm systems and eco-regions;
* Farmers and farmer organizations now understand more about uses of research;
* New research priorities have been identified by scientists/ researchers that take into
account a multi-stakeholder perspective;
* Scientists/ researchers now better understand the problems of application/
* New courses and/or curricula have been established;
* New research-friendly policies, regulations and standards have been established;
* Existing networks are more effective.

Again caution is needed in accepting such consistently positive data without qualifications
but at the same time these results are in agreement with other sources such as case study
material and interviews in NARI.

The difficulty in interpreting these responses stems from the difficulties distinguishing
training and research inputs in terms of their relative effects. An explicit question was
therefore asked to attempt to disentangle reported effects, i.e.: 'How important is training for

sustaining the outcomes from this project and enhancing outcomes from subsequent

Table 7.5 Partner survey: question 21

Relative importance of training (N=140)
Most outcomes are not possible without associated training activities 62 (36.5%)
Most outcomes can be attributed to collaborative research 69 (40.6%)
Difficult to disentangle training/learning from research outcomes 39 (22.9%)

This table suggests that at the very least in the view of a sample of partners, training makes a
significant contribution to the positive outcomes that partner organizations experience.
Again the conclusion that training makes a contribution to Center outcomes and impacts has
been addressed in other studies. However this appears to be the only study that has asked
partners to make this judgement for themselves.

7.4 Country overviews and case studies

Country overviews in seven countries together with specific studies allow for a more
complex and multidimensional representation of the outcomes and impacts of CGIAR
training and learning. Cases incorporate different elements of the NARS including:

* NARI: E.g. Cuu Long Delta Rice Institute Vietnam; INIAP and FORTIPAPA Ecuador;
IRAD Cameroon; PROINPA Bolivia; Department of Agricultural Extension Thailand.
* Universities: E.g. Egerton University Kenya; Universidad Autonoma Gabriel Rene
Moreno Bolivia; University of Dschang Cameroon; Chiang Mai University Thailand.
* Local authorities and other public authorities: E.g. 9 districts within Tien Giang
Province Vietnam; Royal Forest Department Thailand.
* Networks: E.g. CIAT's International Tropical Pastures network; IRRI's Irrigated Rice
Research Consortium.
* Farmers & extension organizations: E.g. Department of Agricultural Extension
Thailand; Extension Services in Mekong Delta Vietnam; Local research committees
(CIALs) Ecuador; Union of producer/processor associations Ecuador.

A summary table of cases and their characteristics is annexed to this report (See Annex IX).
Cases were not selected solely to demonstrate success, but rather the conditions that lead or
do not lead to outcomes and impacts.

The following general messages are supported by these Country based studies:
* In many of the traditional projects where training is significant, i.e., germplasm+new
variety development+participatory breeding+extension work with farmers there are
significant and measurable increases in productivity, production, income and other
benefits to farmers.
* Similarly positive results can be demonstrated in IPM and NRM type initiatives where
training and learning woven into the systematic use of research findings, further
research, controlled experiments, farm-based trials and farmer participatory extension

work can lead to cost reductions, effective strategies for managing plant disease and
improvements in living standards for poor farmers.
* The volumes of training in NARI have been large scale and effective. Many leadership
roles are occupied by CGIAR alumni. Past investments by Centers in training in NARI
can be shown to have led to enhanced capacity to undertake research, changing the role
and relationship of the CGIAR Center to that of colleague and peer making joint
applications to funders rather than providing funding and opening up new research
opportunities for both Center and NARI.
* Many apparently 'local' or 'national' training and learning projects build on Center
experience elsewhere transferring and adapting previous innovations and setting up
methods and models that are themselves transferred and adapted e.g. between LAC
and Asia in Cassava production or across Asia in the case of a rice drum-seeder.
* Training investments in Universities include examples of effective and less effective
capacity strengthening. Different levels of success can be understood partly in terms of
factors that have little to do with the quality of Center inputs, including national
education policy, university leadership and funding availability. Changes in teaching
style/methods and changing methods of selecting students can be especially difficult to
achieve more so than defining new curricula. How funds and training resources are
invested is also important if vulnerable (i.e. non sustainable) 'enclaves' are not to be
* The CGIAR evidently faces distinctive problems in Sub-Saharan Africa, where in some
countries past investments in capacity have not been sustained and NARS are weak.
Whilst these problems are not within the CGIAR's sole mandate, it is seen by
national/regional stakeholders as having a role, together with others. There seems to be
relatively little integration of efforts among the various implicated actors and often
little coordination between CGIAR Centers themselves.
* Sustaining training inputs over extended periods of time seems to be important for
continuity and sustainability. Many successful interventions can be traced back 10 or 15
years, to earlier networks, programs or initiatives. Changes in funding and in national or
local or institutional policies can undermine apparently successful initiatives.
* One capacity result that can be found in a number of cases is policy learning by a NARI
(INIAP in Ecuador, MARD in Vietnam) 'this is seen as the model of for achieving
sustainable agriculture in Vietnam' or government ministry (Ministry of Natural
resources and Environment Thailand).
* Decisions about priorities in a country or region are made for reasons and according to
criteria that relate to the mandates, resources and priorities of particular Centers with
little system-wide overview that might suggest handover to or mobilisation of another
Center. (For example there may no longer be a need to enhance capacity in plant
breeding or NRM but policy and economic or market issues may still be considered
* The shift to in-country training and learning has increased the importance of informal
and innovative teaching and learning methods. There appear to be few resources
available to support or develop or systematise innovative learning approaches.
* There is a strong commitment in many Centers to training, capacity development,
working with national stakeholders and piloting innovation at a national and regional
level. In the words of one senior Center manager: 'Global public goods rest on the
capacities of countries to access and utilize them, otherwise they are not global public

goods.' At the same time there are many results of CGIAR innovation that begin their life
as mainly relevant to a particular national or regional setting and through dissemination
and adaptation usually involving training and participatory approaches become 'global'
at the next stage.

What country studies confirmed and challenged
Field visits to NARS generally came after much preparatory work. It therefore provided an
opportunity to cross-check ideas and sometimes propositions that emerged from Center
visits, early questionnaire returns, pilot investigations and reviews of documentary sources.

Many initial impressions were confirmed by these case studies. For example:
* The difficulty of disentangling training from research. Three configurations were evident:
1) training in order to prepare to undertake research (This would apply to most of the
NARI strengthening examples e.g. Bolivia case 1 Participatory research in PROINPA or
VASI see Vietnam national report); 2) training in order to use available research
knowledge or adapt what is known to local circumstances (See Ecuador Case 2 Cassava
processing); 3) training as part of an ongoing research project (See Thailand Case 1
Participatory mapping).
* The growth of in-country training and learning. The panel encountered training activities
of which there was little detail available in HQ and certainly the volume and types of
training and learning were unknown. The reduction in HQ located training courses was
also noted by interviewees. (See Thailand Case 2 Integrated Cassava Cropping.)
* There was a stand-alone character of some project based and associated training and
learning in-country. Training and informal learning events were entirely the
responsibility of dedicated researchers. There was an absence of pedagogic backup, even
though in some cases there might be a 'framework' or 'guidelines' available.
* The prevalence of informal learning and mixes of different training types tailored to
particular problems and projects. (See for example Bolivia Case 2 Bean production and
Case 1 Participatory research.) Starting from training types conveys very little of the way
a mix of different training and learning modes interact and reinforce each other in situ.
* How relatively small beginnings often when a key individual attends a conference or
training course can lead to major changes in capacity and priorities. (See for example
Bolivia Case 2 Bean Production; and Vietnam Case 3, Enhancing gender equality.)
* The weakness of NARS in some countries and the consequence of not having needs-
analyses and clear priorities coming from the NARS. (See for example Ecuador Case 1
and Malawi and Cameroon national reports). In circumstances where the NARS voice is
weak, the availability of external research funds backing up Center priorities start to
become pre-eminent in determining research with knock-on effects for capacity
strengthening, which are usually asymmetrical more likely to involve teaching in
didactic mode than peer learning.
* The very different kinds of relationships that evolve once a NARS begins to achieve a
degree of capacity and resource. Under these circumstances there is an increased
importance of networks, peer learning and collaboration among 'professional equals'28

28 This and other examples of what happens among weaker NARS are consistent with the framework for
diversified framework for training and learning described in Chapter 2 (section 2.3).

(See for example Thailand national report re. Kasetsart University and Vietnam national
report re. VASI).

On the other hand some propositions were challenged and there were many new lessons
coming from cases of NARS including of NARS partners and specific projects. New
understandings emerged for example with regard to:
* The scale of CGIAR investment in capacity strengthening;
* Continuity and the long term nature of many interventions;
* Preconditions for success beyond the control of Centers;
* Sustainability of outcomes.

Scale of CGIAR investments in capacity strengthening
The Panel encountered many NARI in which the scale and persistence of CGIAR
investments in capacity strengthening was strong. These activities fell into three main

Capacity strengthening at or near start up
* For example IRRI's commitment to Cuu Long Rice Research Institute in the Mekong
Delta in the 1980s and early 1990s; and its more recent work with the emergent NARI
(NAFRI) in Laos would be examples where a Center has made a critical difference and
influenced research agendas, ways of working and openness to international research
networks at a critical stage.

Specific capacity interventions
* For example CIAT project to develop Monitoring and Evaluation capacity in KARI
Kenya (Kenya Case 1) and CIAT's investments in participatory research in LAC see
Bolivia Case 2 and Ecuador Case 2. The contribution of several Centers to the
stabilisation of Bunda College in Malawi, within an otherwise very fragile NARS.

Crisis interventions
* In some cases Centers have taken a leadership role when a NARS was in crisis or close to
collapse. The best documented example is PROINPA Foundation see Bolivia case 3 -
where CIP played a fundamental role in leadership, training and learning over an
extended period, with cooperation from ISNAR at the early stages.

The scale of these investments and their strategic importance cannot be overstated. It is
arguable that in some cases they went beyond the research mandates of the Centers
concerned and focused mainly on the capacity needs of the NARS. However in the Panel's
view these interventions can be justified because they have created or preserved an
infrastructure capable of undertaking future research and sustained partners in key countries
with which Centers can subsequently expand their collaborative research.

Continuity and the long term nature of many interventions
From an HQ perspective it can seem that projects are short term and hence liable to
undermine long term capacity building by reason of their funding. On the ground this
appears not to have been the case in many instances. Exceptionally projects can be long term
when donors have a long term perspective this would be the case with Swiss Development

Corporation's commitment to projects in both Latin America and SSA and Rockefeller
Foundation's long term commitment to CGIAR cooperation in SSA. Projects can also be
'follow-ons' from predecessor projects (the case with ICRAF projects in Thailand). In part
this degree of continuity can be explained by some donor policies; in part it can be explained
by the kind of project profile that appeals to any donor which includes a baseline of
experience, data and personnel.

In addition the continuity that the Panel encountered can also be explained by the personal
longevity in region of key individuals whose personal networks and detailed on-the-ground
knowledge enables them to successfully leverage project funding. This was the case for CIAT
and ICRAF in the greater Mekong Basin countries. This raises the question about how the
Centers maintain and re-create these kinds of strong local and regional networks in the

Preconditions for success beyond the control of Centers
As was noted previously the factors for success and failure are often outside the control of a
Center. Preconditions for success noted in cases include:
* Long term reform in NARI that prepared the ground for a particular intervention (e.g.
KARI case in Kenya);
* National or institutional leadership responsive and able to work in partnership (e.g.
PROINPA in Bolivia);
* A commitment to participatory methods that is written in to the Constitution in Thailand
making it necessary to follow participatory practice including training and learning -
with farmers in all agricultural research and extension work in that country;
* Strong and committed partners whether Universities, NGOs or governments, able to
support Centers, attract funds or take-over what has been initiated (See for example
Universities in Thailand and various NARI in Vietnam).

Whilst these success factors are outside Centers' control they do suggest criteria for future
investments in capacity strengthening and training and learning. On the other hand there are
also factors that have undermined Centers' investments that are also outside their control.
The most obvious example of this among this study's cases is Ecuador Case 2 (Cassava
Processing) which involved a major effort by CIAT in post harvest technologies and
processing and which failed after a period of apparent great success. In this case market
competition (from Thailand), funding withdrawal, natural disasters all reinforced quality
problems and virtually destroyed 17 processor associations and the industry they supported.
On the other hand as is noted below the temporary success of this project has not entirely

Sustainability of outcomes
Despite the continuity of many projects and the continued commitment of many donors and
stakeholders it remains true that some projects end often for good reasons they were
intended to support a specific research project which was completed. It is also the case that
other projects fail in terms of their initial objectives or expectations. This raises legitimate
questions about the contribution that these interventions make to capacity strengthening
whether at individual, institutional or broader levels.

Evidence suggests that it is too narrow a view of sustainability, to conceive it solely as
continuity in the same form of a specific initiative or project. From NARS based cases it is
possible to identify a variety of ways in which training and learning investments linked with
research and other capacity strengthening activities have been sustained. These include:
* Policy leverage: where the project ends but lessons learned are taken up at a policy level
and influence policy innovation. (See case of ILRI smallholders project in Kenya and
Agroforestry in Thailand);
* Institutionalisation: where an institution becomes established and transforms itself by
taking on new mandates and roles. (See PROINPA in Bolivia);
* 'Spill over': Where a single person who received training, mentoring and support can
become the initiator of a significant institutional change process (See Gender equality in
* Replication: where training and learning and joint research enables a partner to replicate a
similar project on its own (See Universities in Thailand);
* Empowerment: where the experience of involvement in an initiative even if its initial
success is not sustained can enthuse and empower individuals perhaps to work in the
agricultural sector or to embark on longer term education and become initiators or
leaders in subsequent agricultural innovations. (See Cassava processing APPY's in

On this basis the outcomes and impacts of Center efforts to undertake research and training
in ways that strengthens capacity in the NARS can be shown to have a far greater impact
than might at first occur.

7.5 Conclusions

The Panel has found strong and consistent evidence of the effectiveness of CGIAR
investments in training and learning often but not always linked closely with research in
strengthening capacity in the NARS. Country based studies in 7 countries and across LAC,
APO and SSA have confirmed impacts for individuals and institutions. The scale of
investments in NARI has been considerable as have been the results. Many of the leaders of
national research in agriculture are Center graduates and the agricultural research agendas
of NARI, government ministries and other NARS partners have been shaped by Center
inputs. In particular CGIAR centers have contributed to the internationalisation of research -
linking even fragile NARS partners to international scientific agendas.

Results of these capacity strengthening initiatives have included modernising and
strengthening NARIs, generating new scientific knowledge, transfer of existing technologies,
the introduction of new crop variants, more effective means of crop protection, sustainable
agricultural practices, increases in farmers' income and increases in productivity and
competitivity of exports. There are positive results in outcomes and impacts.

Country studies and surveys of NARS partners have confirmed the difficulty of separating
out training and learning from research and indeed germplasm distribution. However the
majority of partners who responded to questionnaires and many of those interviewed face to
face confirmed that training was a significant contributor to positive outcomes in
collaborative research projects. These country studies have also confirmed the growing

importance of informal training and learning alongside formal courses. However as
previously discussed many of these efforts are without pedagogic backup or quality
assurance procedures.

Country studies have highlighted the problems that NARS are prioritising and which set the
parameters for many of the interventions and projects in which current training and learning
activities are embedded. These research challenges are often post production, market related,
concerned with environmental problems including drought, seek to work in less favourable
environments with poor farmers and confront policy and regulatory constraints. Given that
many current projects focus on policy development and markets and work with extension
and farmers' organisation, the prevalence of participatory learning approaches and ways of
managing policy dialogues is also understandable.

Contextual factors outside the control of the CGIAR present clear limits to the effectiveness
of its contributions to capacity strengthening. Regional differences were evident in terms of
the likelihood of ex-trainees being able to use what they have learned, a problem often
associated with lack of resource and most strikingly so in SSA. Institutional instability was
also a strong feature of the poorer countries of LAC Bolivia and Ecuador included in the
study. However the success and contribution of CGIAR inputs have been striking even in the
most adverse conditions, especially when working with innovative local partners and
committed donors. The sustainability of the results of past investments in training and
learning increases considerably when account is taken of a broader set of 'results' that go
beyond intentions and objectives. Many projects that have apparently failed have left behind
a large footprint and many investments in training and learning have had unintended but
with hindsight foreseeable positive consequences for NARS.

The serious problems faced by countries where NARS are weak and where Centers in
isolation can only expect to have limited impacts, highlight the need for innovative
approaches to capacity strengthening. These will need to better integrate training and
learning with other capacity strengthening measures and coordinate the plans of more than
one Center together with those of other key stakeholders NARS partners, donors,
governments, and universities.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs