The effect of new marriages among the aged upon the disengagement process

MISSING IMAGE

Material Information

Title:
The effect of new marriages among the aged upon the disengagement process
Physical Description:
xi, 139 leaves : ill. ; 28 cm.
Language:
English
Creator:
Snyder, Paul Wayne, 1932-
Publication Date:

Subjects

Subjects / Keywords:
Old age   ( lcsh )
Older people   ( lcsh )
Marriage   ( lcsh )
Domestic relations   ( lcsh )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 131-137.
General Note:
Manuscript copy.
General Note:
Vita.
Statement of Responsibility:
By Paul Wyane Snyder.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000549725
notis - ACX4022
oclc - 13285702
System ID:
AA00002234:00001

Full Text









The Effect of New Marriages Among the Aged
Upon the Disengagement Process


PAUL WAYNE SNYDER


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DOCTOR OF PHILOSOPHY DEGREE











ACKNOWLEDGMENTS


Tasks are made possible


and worthwhile by purposeful cooperation.


In appreciation of their keen dedication, valuable instruction


constructive crit


cisms,


would


ike, at this time, to record the in-


tellectual debt


First,


owe to several people.


would especially like to express unending gratitude


to Dr. Irving L. Webber of the Department of Sociology of the University

of Florida, for without his extensive knowledge of and vast experience


in gerontology, outstanding courage, and wise counsel


would not now


be entering the realm of scholar


wish a


to gratefully pay tribute to Drs. Felix M. Berardo


and Joseph


Vandiver


, both of the Department of Sociology of the


University of Florida; the former for his professional expertise on


marriage, his


exce


lent guidance, and his depth of understanding; and


the 1


matter for his long-term help and enduring patience.


want to expre


my appreciation to Dr. Harold R. Hunter, of


the Department of


Soci


ology of the University of Florida, for his


assistance and to Dr. James R. Anderson, a member of the Department


of Geography at the University of Florida, for h


Lastly,


Gorman


kind consideration.


would like to thank Drs. Ruth Albrecht and Benjamin


, both of the University of Florida's Sociology Department, for


their very helpful


suggestions.


I I.- I. I i I r .. -t 0














TABLE OF CONTENTS


Pajje


TA BLES . . . ... ...... ..


LIST OF


ABSTPRACT .... ... ....... .............. .... ..... ..... .... ....

CHAPTER


INTRODUCTION. ..... .... ..... .
Nature and Purpose of the Study.
Theoretical Frame of Reference..
Past Research on Older Marriages


. ..a. .t t S S C t *.
..S ..9 .. ... ..4. .. ..* 5 5 .
*. a. .. . ..a.. . S t


METHODOL OGY....
Description of
Development of
Hypotheses.
\Working Hyp
Opera t ional
Instrument


the Sample. ... ..... ... .
Subhypotheses and Working


otheses and Corollaries..
ization of Disengagsement.
Schedules.. ..... ... . .


* S S 4 S S C t S S
* a S P 5 0 0 5


* a P C 4 0 5 4
* a a a U S S S S S
* 9 0 t S S S P a
* P a a t a a a S S S


Analysis of the Marriage Record Data.... .... .....

F I ND I NGS.. .. .. .. ....... .. ... . .. .......

Characteristics of the Research Sam.ples............
Testing of the Central Hypothesis.... ... .. .. .. .
Testing of the Subhypotheses.... ......... .. ..... ..
Testing of the Working Hypotheses and
Coroi laries. . . .......................


FOUr


D I SCUSS ION.. .. .. .. .. .. .. a. .
Age Differential Between the Groups.........
Number of Others in Household .............
Results of a Life Satisfaction Index.......
Tabulations of Responses to Marriage Items..
Nature of Responses to the Marriage Queries.
Significance of This Study on Disengagement
Theory ... .......... ......... .......


S a P 4
* S 5 94 S




. 5 9 a a:
.......

.......


FIVE


PROPOSALS FOR FUTURE STUD I ES OF AGEDNESS
AI D AG:D MAR,. AGES. ....... ... ... ...... ..
r a,, -


TIIREE


A C SN O\!LED Gil EN TS .












TABLE OF CONTENTS (continued)


Page


APPEND ICES


I iNTERV I EW SCHEDULE. . . . .. . ... ...


QUALITATIVE ANALYSIS OF A LIFE SATISFACTION


INDEX: A METHODOLOGICAL NOTE.. ... ...... ....

REFERENCES .. ... ....... .... .... ,. ... .. ... ,

BI O. RAPHICAL SKETCH .... .. .. .......... .. ..... .. .... .. ....














LIST OF TABLES


Table


Page


P~rcetNc


or All


11ra r iayes


in Which the Groom, Bride,


or BothWere 65 Years Old or Older, 1949- 1959........

Percentage of All Marriages in Which the Bride or


Groom We re


Years Old or Older, 1951-1967..........


The Number of All Marriages and the Number and
Percentage of Marriages with at Least One of the
Partners over o60 Recorded in Alachua County,
Florida, 966- 1969, Inclusive. . . ...............


4. Distribution by


of the Partners in Marriages


w th at Least One of the Partners over


Recorded in A1achua County, Florida, 1966-1969,
I nclus ive. ..... .......... .. ... .. a ...... .........


5. Harri~ges in Which at Least


One of the


Pa rtr~ers


Was Over
Be tteean


60 According to


Difference


Bride and Groom, Performed in Alachua


County, Florida, 1966-1969, Inclusive...... ..........

6. Differences in Ages of the Married Pair in the


Cases


Where the Brides Were Older Than the Grooms


in i-arriages in Which at Least One of the Partners


Was over


60, Performed in Alachua County,


Florida, 1966-1969, Inclusive...................

The Educational Levels of the Experimental and


Comp r ison


groups. .... ......... .a .... ........


The Primary Lifetime Occupations of the Expert-


mental and Comparison Groups......


The Present Employment Statuses of the Experi-
mental and Comparison Groups..................


09


r












LIST OF TABLES (continued)


Table

12.


Page


The Total Role Counts of the Experimental and


Compar;- son Grou!os


(Percentages) .... .. .. ... .. ...


The Total Perceived Lifespace


Scores


of the Experi--


rientn 1


and Conparison Groups (Percentages).........


The Lifespace 'i.easure f


Scores


of the Married and


Unmarried Subjects in the Comparison Groups
(Percentages) . . ...... ... .. .. . . .

The Total Role Counts of the Married and Unmarried
Subjects in the Comparison Group (Percentages).......


Total Perceived Lifespace


Scores


of the Harried


and Unmarried Subjects in the Comrparison Group
(Percentages) . . . . . ... .... .. .. ...


The Lifespace MTaasure f


Scores


of the Male and


Female Subjects in the Experimental Group
(Percentages) . . .. . ... ....... .

The Total Role Counts of the Male and Female


Subjects


in the Experin'ental Group (Percentages).....


The Total Perceived Lifespace


Scares


oaf the


Male end


Female


Su~bj acts i


the Experimental


Group (Percentages)......... .. a....................'.


20. Number of Living Children in the Experimental
and Comparison Groups (Percentages) ...................


Total Lifespace ;Measure f


Score


by Number of


L i ving Children for the Exper ixmental and Compari -
son Groups (Percentages) . . ... .. ... ....


Total Role Count by Number of Living Children for
the Experimental and Ccrmparison Groups (Percentages)..


Total Perceived Li espiace


Score


by Number of


Living Children for the Experimental and Comparison
f_ r* lltn n lc rr-s^ni r n^-'-t-r-













LIST OF


TABLES


(cont inucd)


Table


Page


Total Lifespace


I;rsU-e


Living Siblings for


son Groups


Co rbbi ;ed


Score


by Number of


the Experimental and Compar i-
(Percentages)... ..... .. ... .. ....


lTh Total Role Count by the Number of Living
Siblings for Both the Experimental and Comparison


Groups


Combined


(Percentages) ....... .. ... .


The Total Perceived Lifespace


Score


by the Number


of Living Siblings for Both the Experimental and
Comparison Groups Combined (Percentages)..............


Total Lifespace Measure f


Residence


Score


by Rural and Urban


for Both the Experimental and Comparison


Groups Combined


(Percentages). . . . . . .


The Total Role Count by Rural and Urban Residence


for Both the Experimental


and Comparison Groups


Combined (Percentages) . . .. . . . . .


Total Perceived Lifespace


Score


by Rural and Urban


Residence for the Experimental and Comparison Groups
Combined (Percentages)... . .. . . . .. .....


Total Lifespace Measure f


Score


by Age of Spouse for


Both the Experimental and Comparison Groups Combined-
(Percentages)... .... .. ..... ... ... .........


Total Role Count by


of Spouse for the Experimental


and Comparison Groups Separately and Combined
(Percentages) . . . ................ . .....


The Total Perceived Lifespace


Score


Spouse


for Both the Experimental and Comparison Groups
Combined (Percentages)...... . . . ..... ......


The Lifespace lMeasure f


Score


by Race for All Subjects


Thk. Tnt-nail DPlo Cfninra c h'FAl\


(Pe rcen tages)


Fn r










LIST OF TABLES (continued)


Table


Page


The Perceived Lifespace
(Percentages) . ... .


Score


for All Subjects


Ages


of the Respondents for the Experimenta


and Comparison


Nh8 *!:


Groups (Percentages) . . . .


of Others in the Household for the


Experimental and Comparison Groups (Percentages)....

Mood Tone Subtotal of the LSIA for the Experimental


and Compari son Groups (Percentages)


.... .. .. 85


Zest


for Li


Subtotal of the LSIA for the Expert-


mental and Comparison


Groups


(Percentages) . .


41. Congruence Subtotal of the LSIA for the Experi-
mental and Comparison Groups (Percentages)..........


42. Fortitude Subtotal


of the LSIA For the Experimental


and Comparison Groups (Percentage ). ........

43. Self-Concept Subtotal of the LSIA for the Experi-
mental and Comparison Groups (Percentages)..........


Total LSIA


Score


for the Experimental and Compari-


son Groups (Percentages) .. ... . . . . .


The Length of Time the Cooperating Newlyweds Knew


Each Other


Before


They Married (Number and


Percentage).. .... .. .. ... ... .. .. .. .. ...










Abstract of Dissertation


P re 3;n ted


to the


Graduate Council of the University of Florida in Partial Fulfi.llment


of the Requirements for the


Degree


of Doctor of Philosophy


THE EFFECT OF NEW MARRIAGES AMONG THE


AGED UPON THE DI


SEijAGEMENT PROCESS


Paul Snyder

August, 1971


Chai rman:


Dr. Irving Webber


Maj o r Departnen t:


Sociology


This study dealt with the impact of marriage among older persons


upon the disengagement process .


Disengagement theory pos its successful


aging as a process of gradual social and psychological withdrawal with


assoc


iated personal i ty


changes.


The experirint.al population


was com pri sed of persons 60 years of


or over


we r e


wed in Alachua County, Florida, nd whose marriages


w re


1( --


than five years'


duration.


These subjects resided in a


nine-county region i n orith-centra r iorida.


comparative sample of


ol der


icr~ SulS


was selected from the


reg ion


without reaard to their


marital status.


The experimental


samp


consisted of 71 subjects and


the comparison group nurb red 60.

Most of the datawereobtained in the homes of the respondents by


means of an intervi ea schedule administered by the in vestigator.


schedule incorporated,


three


measures of societal engagement:


life-


space measure f score,


a role count, and


a perceived li space score.










background characteristics, and a


the respondents' marital


series


expert enc3s.


of open-ended i teams concerning


Data also were obtained from the


marriage-license applications regarding the experimental population.

The central hypothesis argued that marriages contracted late in


life


serve


to retard the disengagement


process.


Two subhypotheses were


formulated:


female subjects in the experimental group will be


less disengaged than the males in that group


married subjects in


the comparison group will be less disengaged than unmarried subjects

in that group.


working hypotheses and five corollaries


to test the


constructed


influence of number of children, number of sib ings, rural-


urban residc!ence,


spouse,


race.


The corollaries seated that


if statistically significant differences were found for any of the

five variables, the relationship would be more pronounced in the experi-

mental than in the comparison group.


While none of the hypotheses


7i~31: ~


was substantiated on all three


of engagement at an acceptable level of confidence, the data


did lend


scnie


support to the hypotheses dealing with the influence of


being newlywed, being married, having nore children, having more sib-


liings,


I-'-.


g a young


spouse,


and living in a city on delaying the


disengage:ai.nt process.


Likewise the corollaries failed to find support


on al


three


measu res


at an appropriate confidence level.


Neverthe-


less, two of the corollaries, concerned with the relationship of number


of siblings and age of


spouse


to disengagement, received


some


support.


e










of the corollaries, concerned with the number of children,

out in the direction opposite from that predicted.


was borne


The findings of this


study


appear


to underscore the necessity


for continued refinement of the theory of disengagement.


Successful


aging remains a rather elusive


concept


and neither activity nor dis-


engagement should be considered the


sole indicator of it.


measuress of


the aging process such


as the life-satisfaction index may need to be


revised and subjected to further testing on diverse populations.


Future research dealing with the influence of


older marriages


on disengagemerit would profit from use of the longitudinal


deijgn.


Con-


prison of similarities and differences between marriages contracted at


younger and older


ag-s


could extend our knowledge in the fields of both


social gerontology and sociology of the family.








*u












CHAPTER ONE

INTRODUCT I ON



Nature and Purpose of the Study


In this country, and some others, both the number and the propor-


tion of older people in the population have been rising.


"This increase


is due to a variety of factors.


tat ion


Such things


"miracle" drugs, better diets, new surg


spread health programs


keeping more peop


and higher leave


improvements In sani-


techniques, wide-


of living have succeeded in


alive all along the age continuum.


When such develop-


ments are coupled with a reduction in the birth rate and low leve


immigration


, the population begins to age.


With more people reaching


older


ages


and with fewer new members being admitted


into


a society,


the population pyramid begins to bulk large at the


higher


ages.


The results of these trends are reflected in the following data:

in 1940, persons aged 65 and over in the United States numbered

9,019,314 and comprised 6.7 percent of the population; in 1970, the

corresponding figures were 20,049,592 persons and 9.87 percent of the


population (U.


Bureau of the Census, 1970:4).


This dramatic demographic change has generated a high degree of


interdisciplinary interest


n the relatively new field of


Soci


gerontology.











A review of the


existing literature reveals that most of the


research on the problems of the aged has dealt with death, bereave-


ment, health, widowhood, physical infirmi t ies


, nursing homes


adjust-


ment to retirement, and other problems typically encountered by very


old persons.


It is clear that the ordinary or expected phenomena


concerning our senior citizens have received increased attention from


soc la


scientists.


With the greater part of the research being focused on decre-


rents in health,


oss


of roles,


et cetera,


little attention has been


given to the resumption of old roles or the assumption of new roles,


especially those implied by new marriages and remarriages.


exists


This


despite the fact that new marriages among the aged,


as will be subsequently shown


, are relatively common and important


This dissertation will focus on elderly newlyweds and the effect


that new


marriages


contracted by aged persons have upon the disengage-


ne n t


process.


T heo reti


Frame of Reference


Cumming, Dean, Howell, and NcCaffrey formulated the theoretical

framework describing the disengagement process (1960:14-23).


This process is most often thought of


as a mutual withdrawal


which


takes


ace


between aging persons and others in the social


system.


several I


dimensions of the process, along with illustrations, will


1,0 n roe on ii r t-.r,. na~ I- el, ~r.4 at- r-n 4- I-.,,. r a nr r..~r 4-c ~ rn nnn r:~ 4-7 ann 1 : ~












mutual withdrawal, resulting in reduced interaction between the aging


person and others in the social


system.


An old person becomes less involved in the life around him than


when he


was


younger, and his withdrawal may be accompanied by an in-


creased preoccupation with self.


The p


process


seen


as function


for soci


Therefore


is a society


expectation, and


reciprocity is implicit in it.


Indeed,


the social order must have


some


means


that pave the way or make it


easy


for the elderly to move toward eventual disengagement, and the older


person is expected voluntarily to


cooperate


in the process.


When the process is complete, the equilibrium which existed at


middle age between the individual


and society has given way to a new


equilibrium characterized by


a much


greater


distance between the aged


individual and his society.


Three orders of change should be distinguishable


e in disengage-


ment.


First,


there


should be an observable reduction in the absolute


number of people with whom the aged person interacts.


Second, there


should be


a qualitative change


even


in the interaction that remains,


especially in terms of time and purpose.


changes.


The type of activities actually


The aged person reduces activities that involve the outside


world and takes on activities that


are more


inner-d i reacted.


He gives


up activities that call for interaction with others and adds those


, it













activities that center on self (Cumming and Henry, 19(1:8-9 and 38-45).


Lastly,


we should


see some


patterns of personality altering to


accom-


modate this increased involvement with self.


The last is


seen


tsa


circular procedure that both


changes.


causes


The older person is conceived


is the result of the first two

as being at the center of a


network of social interactions and,


as he


ages,


his social life con-


stricts and he experiences a curtailment of involvement in the social


system.


Disengagement is viewed


as a process which is beneficial to


both the individual and


society.


There were three basic points in the original work by Cumming


and associates.


They stated that the process of withdrawal is modal


for the aged, that disengagement is both intrinsic and inevitable, and


that undergoing the p


process


is a condition of successful aging.


(jeniys, in a later work, redefined the disengagement process

as not necessarily inevitable but still maintained that the equilibrium

that existed in middle life between the individual and society gives

way gradually to a new equilibrium characterized by greater distance


(196


-S-SI'


-393).


Williams and Wirths (1965:3) agreed that the process is intrinsic


to aging but probably not inevitable.


best defined


They stated that disengagement is


as a process of mutual severing of ties between the indi-


vidual and the social milieu in which he li


ves.


process


involves




5





a universal, though it may vary in degree of graduality from one culture

to another and it may not be inevitable for all people in each society.

Two research projects (Thompson and Streib, 1961:200; and Maddox,


1965:117-130) even questioned that


o i sengagem'ent


was


a mark of


success-


ful aging.


Maddox concluded that disengagement,


as a correlate of aging,


found much more frequently in the very advanced age levels--a time


when physical and mental decline i's most likely.


Thompson and Streib


also found that disengagement worked only for "very old"'


categories s;


by "very old" they meant 80


years


old or older.


Two articles (Kapnick, Goodman


, and Cornwell, 1968:305-310; and


Glenn and Grimes, 1968:563-575) clearly demonstrated that not all social


involvement decreases with


age.


The authors proved that political


interests and activities in


crease


wi th


aye.


Videbeck and Knox (


965:37-48)


challenged the modality of dis-


engageene~


In a longitudinal study,


a part of which


measu red


involvement level of the subjects, they found that 90 percent,of the


over-65 nonparticipants


were


also nonparticipants five


years


earlier


and 90 percent of the 65-year-old group and older who


we re


participants


were classified


as participants five


years


earlier.


The participation


variables employed were church and club activities, public meetings

attended, time spent -reading, and political activity. This finding


seems


to indicate that a person's adjustment in the latter part of


his life-cycle will merely reflect


a continuance of


a life style


was













Although the evidence presented


above


'JI.t-2


appear


to throw


doubt


on some


of the basic tenets of the disengagement theory, there


does


seem


to be general


agreement


that the process, at least,


does


exist.


Most


social


scientists


are even


will i n


to grant the mutual it


or disen


ment


Certainly, given


a general


lied


natural reduction in


energy


levels,


aged individuals should make willing accomplices in the


process


of their


separate ion from


society


i nvo 1 vemen t.


Even the opponents


cI dsna


ment theory accept the direction of the


sociocu


tural


forces


which


generally against


a high leave


of participation in


society


by the


elderly


Hence, an intellectual challenge


is clearly


present


to empirically


define this mutual withdrawal,


means


ure to what


extent it


is occurring,


and outline situations (other than increased aging) that


ei ther


speed


up or retard the process.

Havighurst (1961:8-13) posited the disengagement theory against


the activity theory


The activity theory


developed through the


years


by Cavan (1949:71-83)


Beard (1949:274-279), Albrecht (1951:380-386)


Ph ili ps (1 931:215-2


Burgess (1960:271-298), and Will


laes,


et. al. (1963:261)


hypothesized that a high level of activity and social involvement is


more


desirable adaptation for aged people to make in


sac iety,


Hav ghurst stated that the activity theory is favored by the symbolic

interactionists and social practitioners who suggest that people should


-,C~


are











so doing become the passive victims of society.


On the other hand, the


disengagement theory is favored by functionalist theoreticians who view


it as constructive and

addiLion, they see the


necesso ry


aged


for the continuity of society.


as cooperative actors in the process.


Therefore, the disengagement theorists maintain that dis


engage-


ment reflects successful aging, while the activity theorists claim that

activity represents successful aging.


From the above arguments,


it looks


if the two positions are


indeed quite irreconcilable.


But,


as a matter of fact, they are


exactly


reciprocal.


The results of


one are conceived in the degree of social


activity and the consequences of the other are conceived in the degree

of social isolation.


In real ity


, then, a test of one of the theories automatically


becomes


a test of the other.


This study will do both, although, for


methodological reasons only, the effect of the new unions wLi


stated


as being measured against disengagement.


Past Research on Older Marriaqes


A thorough search revealed that few research projects have


been undertaken with aged newlyweds


ever


the subject of investigation.


One study (Bernard, 1956:50-66) examined remarriage in all


age groups,


hence was not primarily focused on new marriages among the aged;


however, statistics were gathered for advanced


age groups which should












That the proportion of all marriages that


are remarriages


Increases


wi th


because


the proportion of persons


who have


been


previously married


increases


with


age.


That most


the previous marriages in the young to early-


middle-age groupings ended in divorce while the majority

of those in the late-middle to old-age groupings ended in

death.


That approximately one-third of the women


remarry marry younger men while only about


over


percent


the men over 6


who remarry do


An important work done in the


area


so with older

is reported in


women.

a University of


Connecticut monograph (McKain:1969).


However


, the major thrust of that


investigation


was


to ascertain the marital adjustment of older peop


were


marrying again in their later


years


and to develop a


scale


to measure the


success


potential of


a retirement


remarriage.


The McKain study consisted of


a samp


of 100 couples,-; with no


control group.


He included only those unions in which both partners


had been married previously.


Only brides who were 60


years


old or


ove r


and grooms who


were


at least


years


of age at the time of the


wedding


were


considered for


section in the study.


Both the Bernard and McKain studies had their basis in the


testing of theory in the field of marriage and the family and they


were


nfl rn T; lx. cr n n rhar r; ~n Iei h, rn r-cr t~InciA td-~~~ +kr nnr ki mm f mt it:n nf tkc r











no research to date with aged newlyweds


in a


as subjects and based


gerontological frame of reference.


It is true, however, that in the McKain study (1969:101) the

effect of the remarriages on the process of disengagement was inci-


dentally tested.


The data


seemed


to indicate that marriages in the


later years counter the disengagement process in familial relationships,


but the reduction of other roles in society continues.


However, the


results were inconclusive.

Both the Bernard and McKain studies excluded aged subjects who


were marrying for the first time.


It is true that never-married sub-


jects would have no relevance in studies seeking to describe the


keys


to successful remarriage.


However


, even aged bachelors and spinsters


becoming newlyweds have to


cope


with disengagement,


so they will not


uded from the present investigate ion.


McKain (1969:1-13) strongly contended that there is a societal


ethic


against


elderly people marrying and that the children of the


elderly


are very much opposed to the remarriage of their parents.


However, the mimeographed data tables that his work was based on clearly

indicate that the majority of the children of both the brides and the

grooms encouraged the marriage of their parents (McKain's Retirement


Marriage Statistica


Tables, 1969:4-5).


Moreover, Bernard (1956:25)


suggested that throughout society

of remarriage in all aqe groups.


.y there exists an institutionalizing

One reason that more older people


exc


I











is greater that that of those of the past.


The advances in medicine


have


done


much more than create an increase In life


expectancy.


drugs and surgical techniques have brought about better health and im-


proved physical vigor


among


the older population in addition to extend-


ing their lTfe-span (Rose and Peterson,


965 :364;Youmans,


:201)


Hence, with a prolonged period of energy and potent


usefu


ness


, many old people in


a sense


are reverting to an earlier


stage


the life cycle and becoming newly


yweds.


The data in Table 1, accumu-


lated by the Administration on Aging, attest that new


marriages


con-


traced by older persons


are


uncommon.


(The source does not give


the absol


ute number of marriages.)


Table


Percentage of All Marriages in Which the Groom, Bride, or


Both Were 65


Years


Old or Old


, 1949-1959"


Percentage of All Marriages in Which


Year


the Groom, Bride, or Both
65 or Older


were


1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959











Table


also indicates that marriages among the elderly


were


the increase during the


years


shown.


Up-to-date data


were


sought to


ascertain whether or not this trend continued.


Exactly comparable data


were not available, but by using information contained in the Vital


Statistics of the United


States


(1960:1-31 and 1968:1-13), it


determined (Table 2) that the trend toward increasing incidence of


marraige among the elderly


over


time did not continue beyond 1960.


Table


Percentage of All Marriages in Which the Bride or Groom


Were 65 Y


ears


Old or Older,


951-1967'


Percentage of All Marriages in Which
Year the Bride or Groom were 65 Years
Old or Older

Bride Groom

1951 0.6 1.5
1952 0.6 1.6
1953 0.7 1.7
1954 0.7 1.8
1955 0.7 1.7
1956 0.7 1.8
1957 0.8 1.9
1958 0.8 2.0
1959 0.9 2.0
1960 0.9 2.0
. 1961 0.7 1.8
1962 0.8 2.1
1963 0.8 1.9
1964 0.8 1.9
1965 0.9 2.1
1966 0.8 1.8
1967 0.8 1.7


was


__ L


--












Table


shows


a tendency toward increase up until


1960, then


a plateau appears between 1960 and 1965


and it


seems


that an actual


decrease has occurred


since


965.


The U.


S. Vita


Statistics (1964:


I-18,


965:1-17, 1966o:1-14, and 1967:1-14) did give the absolute


number of aged marriages


for the latest four years for which data


available.


In 1964


, 10,831 women


ove r


65 got married and 24,920 men


over 65 got married


, for a total of 35,751.


In 1965


, the corresponding


figures were


1966, it


was


2,248 women and


12,430 women and


,328


men for a total of 39,576.


,115 men for a


total.


1967, it


wa s


12.401


women


and 26


,537


men for


a 38


,938


total.


Both the disengagement and activity theorists point out the

gradual role loss of older people in their continually contracting


social world.


Terms such


as role changes, loss of roles, role


role, rolelessness, and role shrinkage


of gerontological


are commonplace in reports


surveys.


The number of role


a person has is dependent upon the definable


positions, or statuses, that he occupies in society.

whether it is allowed that each status may have a mu


s is true


iltiplicity of


roles or whether it is given that all the usual ro


tasks


assi


gned


to a particular status


are considered


as one rol


Either position


can be meaningful


when trying to correlate role loss with disengage-


ment.


A man who was


a father at age 30 wi ll not lose that status


-J -. --


are


ess












But in this study, such discrimination will not be


necessary,


for ;.n aged person who gets married


assumes


a new status, that of


spouse, and, thereby, by definition


assumes


a new and additional


role (or roles)


aSSI


gned to that


status.


This role gain, then,


theoretically might tend to hold off disengagement.


Moreover, an individual


world relevant to him,


life space


or the extent of the


associated with disengagement.


Back and Kenneth Gergen (in Simpson and lMcKinney


, 1966


Kurt


:289-293)


point. out that an individual who marries may effectively double his


life


space.


Al] his mate's relatives, friends, interests, possessions,


habits, and hobbies suddenly impinge upon his environment.


Therefore,


this interference in a formerly constricting life


space


should slow


down disen


gageaent.


Hence, the major hypothesis to


be tested in


this work is:

New marriages contracted by the elderly tend to have


a braKing effect on their disengagement from


society


Subhypotheses, working hypotheses, and corollaries will be


derived from this central hypothesis.


They


are presented in the


following chapter and


are based on the common demographic variables


such


as maritaT status,


sex,


race,


residence, number of siblings,


number of children, and age of spouse.

hypotheses will be described in detail


The method of testing these

in the subsequent chapter.













CHAPTER T!0O

METHODOLOGY


Description of the Sample


The experimental universe in this investigation is comprised

of those marriages in which either or both partners had attained the


age of 60 at the time of the wedding.


from viewing disengagement


defin i tion follows


as a process that some people begin earlier


than others.


Furthermore


, age


60 seemed more realistic than 65


because


so many people retire now on pension plans before reaching their


sixty-fifth birthday and many others under 65 are retired from active


work due to disabilities.


Moreover, a process, to be adequately


studied, should be examined in its


earl


as we 11


late phases.


The universe was further defined


as all


those marriages with old


persons

Florida.


as partners that were

A county setting was


recently contracted in Alachua County,

selected because the county is the


registration unit for marriage licenses.


However, it has added


advantages in that its clearly defined political boundaries, which


can be precisely located on


a map, enable the researcher to delimit


a specific geosocial area for experimental purposes.


The definition


I
Of course


this


argument could be carried to the extreme by


affirming, and correctly
that moment on a course i


that aging begins at conception and from


-3


charted toward the ultimate end of life




15





of a new marriage, for the purpose of this study, was one that has been


in existence for


aSS


than five years.


In addition, it


was decided


that the marriage should be at least si


some


months old to insure that


time for adjustment had intervened between the wedding ceremony


and the interview.


Since the interviewing took place during the


summer of 1970, the optimal choice was al


at the courthouse


those marriages recorded


in Alachua County, Florida, from 1966 to 1969,


inclusive, who had passed their sixtieth birthday at the time of the


marriage and were


giving in Alachua County or in counties that are


contiguous with Alachua County.


parsimony


This plan not only represented


as regards time, money, and effort, but it also should


yield the maximum usable results since this nine-county "ridge

section" of north-central Florida is relatively homogeneous in terms

of geography, economics, and culture.

This investigation did not include common-law marriages, even


though


some


of our aged citizens are cohabitating under such


ci rcum-


stances


(Wikler and Grey, 1968).


Wikler and Grey maintained that


prior to 1965 a widow who remarried lost all her


soci


security


benefits from her deceased husband and that was the salient reason


why many senior citizens did not legitimatize their unions.


These


authors reported that,mainly through their efforts, the federal law

was changed in 1965 and that now an elderly person can remarry with-

out losing any of the benefits from the covered employment of a











Since the data already presented in this study did


not show


an-increase in formally officiated marriage ceremonies


since


1965,


but instead indicated a slight


decrease


in the trend,


the researcher


checked the present Old


Age,


Survivors, and Disability Insurance


regulations.

According to the information contained in the Social Security


Claims Manual


(July


, 1970:410-420), the termination of widowhood can


result in financially adverse situations in the following


A widow under 60 who remarries


loses all of her


A widow under 60


cases:


a man of any


social security benefits.

, with dependent children in her


care, loses all her share of the OASDI grant upon


remarriage to a partner of any


age.


A beneficiary,


because


he or she was married at


least 20


years


a now deceased worker before


they were divorced,


loses all social security


benefits upon remarriage, regardless of age.


A widow who remarries between the


ages


of 60 and


62 has her
reduced by
man of any


social security check automatical
32 1/2 percent upon remarriage to


age.


A widow who remarries after a
OASDI check automatically cut


2 still has her
1/2 percent, bu


if she marries


a man


over


65, she becomes eligibi


for 50 percent of his award.


However, if 50


percent of her present husband's check is


than


percent of her deceased spouse


s check


or if she marries a nonrecipient of social security,


there will be


a net loss of income.


This does not take into account the possibly similar negative

circumstances that might take place under the many private pension


ess











funds, or labor union plans when a surviving beneficiary remarries.


It al


does not take into account the possibility of


a decedent's


will being worded


so that the surviving spouse is made the trustee


of the


estate


until remarriage with control then going to the


eldest


offspring.


Needless to


say,


there yet remains amp


e incentive for


many older people not to have their present unions officially known


which brings us to the main methodologica

common-law marriages from this study. Fi


reasons


for excluding


rst, since there


are no


records, there can be no accurate count, thereby making it impossible

to reduce the potential for sampling error, especially in terms of


representat i veness.


Secondly,


a field worker interviewing in


business suit and armed with


a clipboard and recording pen would


present a very clear threat to people living under this condition,


which would make effect


cooperation extremely doubtful.


Since the investigation is an experimental design aimed at


discerning differentials in the nature and rate of disengagement,


comparTson sample


was


used.


This


was accomplished by


cons


i during


the nearest neighbor


over


as a comparison-group subject.


experimental subjects


were


asked to


indicate the house


nearest


theirs which had someone 60


years


or older residing therein.


In those


cases


where such households contained two people over 60,


both were interviewed


as comparison subjects.


with elderly people living in them


were


If two or more homes


equidistant from the home of












north was selected, or if there


was none in that direction, the


nearest to the south


was chosen


, or if none there, then


west,


so on.


Development of Subhypotheses and Working Hypotheses


The main hypothesis predicted that the experimental group will

experience a lesser degree of disengagement than the comparison group.


Two subhypotheses were postulated from this.


Since the comparison


group


was


of peop


selected without regard to marital


are


state, it will consist


married, single, divorced, and widowed, and it is


very likely that many subj


ects


in that group will be unmarried.


And,


since the central assumption of this investigation is that newly


contracted marriages have


a negative catalytic effect on the disen-


gagement process, then it follows that married aged people in general

will be less disengaged than unmarried aged people.


Subhypothesis A:


Married subject
group will be 1


unmarried subj


s in the comparison


ess


ects


disengaged than
in that group.


Since the experimental group will be for the most part living


in intact marriages,


age-60 spouses of both


and, therefore, it is expected tha


sexes


many over-


will be in this group, another subhy-


pothesis can be formulated.


The most important crystallization point


of disengagement for men is retirement, whereas for women it is


widowhood (Cumming,


:378-389,


and Berardo,


970:11-25).











Remarriage for


a woman negates widowhood, while remarriage of an


elderly man has little or no effect on his retirement.


Subhypothesis B:


Female subject
group will be


subj


ects


s in the experiment


ess


disengaged than


n that group


Working Hypotheses and Corollaries


Each working hypothesis will be accompanied by a corollary.

The working hypothesis will be tested with data on the combined sub-

jects i.n both the experimental and comparison groups together and

the corollaries will be tested by noting the strength of relationship

in each. group when and if the variable being tested by the working


hypothesis produced a

Cumming et al.


gnificant difference over both groups. 3


(1960:25-32) postulated that disengagement


increases


as the


size


of life space lessen


and the amount of inter-


act ion


decreases.


Hence it can be deduced that


, in general, if an


aged person's family of orientation or family of procreation are

small, or, if there are fewer people in the immediate area of the


aged individual, then both his life space and potential


for inter-


action are diminished.


Working Hypothesis 1:


The larger the number of living


children an aged person has


, the less


effective


I be the d


engagement


process


3The reason al


the hypotheses


were


not merely numbered


is that


there is a fundamental difference between the central


hypothes


s, a sub-


--












Corollary:


If there is a significant difference
noted due to the number of living
children, the relationship will be
stronger in the experimental group
than in the comparison group.


Working Hypothes is


larger


the number of living


siblings an aged person has, the


less effective will


be the d


sengage-


ment


process.


Corollary:


If there


a significant difference


noted due to the number


Living


siblings, the relationship will be


strong


r in the experimental group


than in the comparison group


Aged citizens livi ng in sparsely populated


areas


should also


have smaller life


spaces


than


those


living in cities.


A dense popu-


lat ion,


n addition, should


mean


a greater concentration of activities


and organizations servicing the


aged


and Rose (1963:131) found that


urban elderly have a greater number of friends and acquaintances


than rura


elderly.


Working Hypothesis


The urban aged will be less dis-
engaged than the rural aged,4


Corollary:


If there is


a significant difference


noted due to rural or urban resi-
dence, the relationship will be
stronger in the experimental group
than in the comparison.


Since


years


some


age,


of the spouses


seems


n both groups will be under 60


reasonable to assume, at least hypothetically,


rip'-',


4Af


subjects who


DT 1 -4-,~~~Ar1 ar8


reside in the city limits of Ocala


_ era


Lake


... I


I a i ar A I r -~ r *'.e* -Ir ,r u-t-n


flnr rl?


!











that an older person married to such a younger mate might be in a


situation that would act toward the postponement of the


older spouse's


disengagement.5


Working Hypothesis 4:


Among the married subjects, those


whose
subj


mates are too young to


ects


will be less disengaged


than those subjects whose marriage
partners are also subjects.


Corollary:


If there is a significant difference


noted due to


age of spou


relationship will be stronger in the


exper


mental group than in the compari-


son group


Race may also be


a factor in the disengagement rate of senior


citizens.


Certainly, at


age 60, Negroes, in general,


are closer to


the end of their lives than Caucasians, due to a differential in


their life expectancy.


The a


average


number of yea


rs of life remaining


at age 60 for whites was 18.1 in 1956 compared with 16.5 for nonwhites


(Vita


Statistics of the United States, Mortality Tables,


966:4-5)


A differential of 1.6


years


may not


seem


great, but it is a


measurable and factual difference.

and important demographic variable


Moreover, race is both


hence, it would be wi


a common

to test


a racial hypothesis in order to determine


f this characteristic makes


a difference in degree of disengagement.


these marriage


study did not endeavor to focus on the sexual
es. Gerontologists know, of course, that sex


aspects of
quality goes


on as


long


life goes on (Wikler and Grey, 1968, and Rubin, 1970),


but the frequency, the problem of birth control techniques,


and many


. A- L. -


fl~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~r *!Sj urtFn cfr 1V COYAnrc


rvf mnnrtt


rl n~rCP1


'












Working Hypothesis


Blacks will be more disengaged than


whites.


Corol 1


ary:


If there


a significant difference


noted due to


race


, the relationship


will be


stronger in the


expe


rimental


group than in the


comparison


group.


This makes


a tota


of'


hypotheses to be tested in this work:


the major or theoretical hypot


is, two subhypotheses, five working


hypotheses, and five corollaries.


Operationaltzation of Disengagement


The single dependent variable in this treatise is the disen-


gagement process.


The most concrete terminology


seems


to be in con-


ceptualizing disengagement


as social


isolation


vis-a-vi s


activity.


Cumming


et al .(1960:33) maintained that the concomitant


ements of disengagement


are


ess


involvement with work, family,


and clubs, but suggested that there may be more involvement in

recreation, especially at the early stages of disengagement.

One could argue that. in the early stages of disengagement,


there may be more involvement with fam


y, friends, and voluntary


associate


ions, because retired


ndividua


would have more free time


for such activities.

reduction of familial


Nevertheless, it seems

interaction begins even


more likely that the

before the elderly


individual stops working.


The involvement was more probably most


intense during hi


early-middle


when his children


were


living


soci













age when his married children


chi ldren..


were


The later interaction


presenting him with


processes


new grand-


more than likely involved


a gradua


and mutual wi thdrawal.


As regards


friends, Blau (1961:429-439) showed that the extent


of friendship participation


decreases


with


age.


Turning to voluntary associations, the membership of voluntary


associations, like the population of a society, would not like to


see


their ongoing work and activities disrupted.


A death of


someone


position of responsibility for such continuity would be upsetting to


system.


To avoid this, the individual should ready himself for


total disengagement and the group


expects


him to do


Hence, it is


more logical that a measure of disengagement would be an aged person's


giving up his club membership, and especially yielding any ro


have


he may


as an officer in the group.


Cummi ng, in


a later work with Henry (1963:15 and 51), reaffirmed


the earlier disengagement definition (Cumming et al., 1960:25-32) that

an aged person begins an absorption with self and becomes less and less


involved with


ife around him.


However, the authors concluded that,


since


an older person


sees


fewer people,


less


often, and for shorter time


periods, it is conceivable that


even


their recreational activities begin


to diminish.

Zborowski (1962:300-305) reported that participant recreational


a




24





Physically, of course, it is anticipated that old people will

naturally be limited in the types and amount of recreation pursued,


especially


as regards highly active dancing, athletic


etc.


However,


disengagement should result in a reduction of participation even in the


things they


can do, and, most especially,


games,


hobbies, and pastimes


that require that other people take part.


It can be argued that


since


disengagement is


a preparatory


process for death, it might


seemn


likely that the elderly would manifest


increased church attendan


readying themselves spiritually in order


to die in a state of religious

(1964:149-150) held that the v


grace.


ery


However, Neugarten and


old manifest reduced soca


Associates

interaction


in all


area s


of life.


They begin demonstrating an increased preoccupa-


tion with self and


a decreased time and


emotional


investment with others.


Cumming and Henry (1961:91) also concluded that religious interests


decrease in older


age.


Therefore, the present investigation


assuLimes


that, if an older person is remaining active in the totality.of religiosity
4
by still attending church regularly, teaching Sunday school or catechism

classes, or the like, then he is opposing disengagement.


Cumming et al. (1


590:25-35)


and Neugarten and Associates (1954:


149-150) found that older people are more carefree in their persona


actions and less evaluative of what other peop


think of their.-behavior.-


As a matter of fact, most older people


seem


to abstain from intense


interpersonal relationships.


This produces a nonnormative patterning











reduce the situations in which others exert normative control over him.


The result is


a more


self-centered and idiosyncratic style of living.


One operational


measure


of disengagement would be the degree


of withdrawal from the dimensions of social involvement and whether


the subject


percei ves


such noni nvol vement.


The elements of social


involvement that


were


examined in this study are:


1. Work,


Fam ily,


. Friends, 4. Clubs, 5. Church


6. Number


of others in household,


. Neighbors,


Interaction with specific people, 9. A role count


based on these dimensions, and 10. A perceived life


which asks the respondent


space


to compare his present level


measure


of social


involvement with relat


ives,


friends


, church,


ubs, and people in


general with his level of such acti


ies when he


was


forty-five


years


old.


Instrument Schedules


Two instruments of long standing in gerontology


, each pre-


viously validated in the field,


were


uti 1 ized.


One is the measurements


of engagement developed by Cumming and Henry and found in Appendix


of their book, Growing Old (1961:243-256).


exact


scales


used are


entitled:


"Lifespace


Measure," "Church and Voluntary Organization


Participation," "Role Count," and "Perceived Lifespace."


the "Life Satisfaction Index A" (L


and Tobin (Havighurst, 1961:312-3


The other is


A) developed by Neugarten, Havighurst,


15). The purpose of the latter tool is











does not mean that there is no connection between engagement and life


satisfaction.


need, in the accompanying tables of the McKain mono-


graph (1969:7-8) the majority of the


aged


newlyweds


scored


in the high-


est bracket


life satisfaction


But in another study of older people


in general


CMaddox,


964:121-122), it


was


shown that life satisfaction


does not change with


decreasing activity with


increasing age,


age.


even


At any rate


in subjects who demonstrate

, the determination of such


a relationship is outside the purview of th


study.


But if a signifi-


cant difference is noted on the variable disengagement


amor,


g the


subj


ects


and this is correlated with the results of the LSIA, then this


would act


as a test of the relationship.


instrument also contained additional items defining the in-


dependent variables and persona


background (see Appendix).


In addition,


seven


open-ended probes aimed at eliciting


some


information about their


marriage


were


asked of the newly married couples who indicated that


they would not mind answering them.


serve


The purpose of these items


was


as a guide for future research rather than test any of the hypoth-


eses.


same


schedule


was


administered to the respondents in both


the experimental and comparison groups.


In those


cases


where both


spouses


were over o0, the husband and wife


were


interviewed during the same visit,


but separately.


However, the


seven


queries on marriage were administered


together to couples who were willing to respond to them; if their re-












Analysis of the Marriage Record Data


A total of 97 marriages were


egitimized in the nine-county


area wherein either the bride or groom or both were 60 years of age or


over, during the time span indicated.


Some useful


information was


already on the marriage license applications.


These data include those


persons who could not later be located to be included in the experi-


mental


population.


Data were also found on both marital


partners


which included age, race, community of residence at the time of appli-


cation, previous marital


status, number of previous marriages, how the


last union ended, and if the marriage just previous was terminated by

divorce, when and where.

The number and percentage of those formalized unions are pre-


sented in Tabl


The figures from Alachua County, Florida, depicted


The Number of a]


Marri a


ges and the Number and Percentage of


Marri


ages


with


at Least One of the Partners over 60 Recorded


in Aiachua County, Florida,


1966-1969, Inclusive


Number Number and Percentage of Marriages with
Year of at Least One of the Partners over 60
__Marriages .Number Percentage

1966 901 25 2.8
1967 924 18 1.9
1968 1,114 30 2.7
1969 1,170 24 2.1
Total 4,109 97 2.4





28






in Table 3 confirm the evaluation that the incidence of aged marriages


seems to have reached


S orie


sort of erratic plateau and that the decrease


since 1966 suggested in Table


probably


was only a slight rill in


that plateau.


The analysis of the applications yielded


Soaifl


other interesting


characteristic


of the population.


First, 63 or 65 percent of the couples


were


white, and


4 or


35 percent of the coup


we re


black.


There


we r e


no interraci


marriages among these aged newlyweds.


Of the 194 partners in these weddings


180 or


percent had


been married before.


these, 70 percent had only one previous marriage;


18 percent had two marriages before this one;


marriages previously,


percent had three


1/2 percent were starting their fifth marriage


at this time; and 1/2 percent (1


case)


was entering his sixth marriage.


More surprising is that 7 percent (14 cases) of the population had


never


been


married before.


None of the never-married grooms was uniting


wi th an


of the never-married brides.


It would


seem


because of our definition that only one of the


mates need be over 60 that, perhaps,

be partners who had not yet attained


most of the never-marrieds would

age 60; but, In fact, 8 of the


14 never-married-before group


we re


years of age or older.


Logic might also lead us to suspect that the majority of the


^ .. .... ... ..... 1J L ? L ? J-- L .. .... ... r. r-_ 0 -r












the never-married-before groups who were under 60


years


were


grooms.

Reason might point us toward the assumption that Negroes, due

to their increased chance of being in a previous unrecorded common-law


marra ,;


would compose most of the officially never-married-before


group; however, Negroes accounted for only half of that group.


these


seven


Negroes, five were


over


60 and two were under 60, three


were


rooms and four


I'b1 c


brides.


No Negro grooms


were


under 60,


while two of the Negro brides


We -'e


over 60 in the never-married-before


group.


Of the 180 subjects who did have previous marriages, their


most recent unions were ended by death


19 times, by divorce 60 times,


and only one subject terminated his last marriage by annulment.


is not one of the 14 never-marrieds discussed in the foregoing


graphs because he had been married twice before.)


ment with Barnard (1956:50-66)


among


para-


This is in agree-


found that most previous marraiges


the older age groups ended in widowhood.


The surprisingly broad age

marriages can be noted in Table 4.


range of persons involved in these


The modal age bracket was


60-to-64-year-old group with the tota


distribution being


htly


skewed


toward the older


age groupings with 71


cases


falling in the


65 or older categories and


cases


being under the


of 60.


Tremendous variation existed in the relative


Saces


of those who
















Table


stribution by


of the Partners in Marriages


with


Least


One of the Partners over


60 Recorded


in Alachua County, Florida, 1966-1969,
Inclusive


Number of


Group


25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90-94

Sum












Table 5


Marriages in Which at Least On


of the Partners Was


over Ag
Performed


CCc


rding to Age


achua County


Differ


Florida


ence


Between


1966-1969,


Inclu


s ive


Age Diff. Number Age Diff. Number Age Diff. Number
in Years of Cases in Years of Cases in Years of Cases


the marriage contrasted sharply with the three couples whose age


differential


was 30


years


or more.


The youngest subject in the


roup


wa s


a 27-year-old fema


married to


a 65-


year-o


d male; this union also accounted for the widest


gap (38 years).


united to


A 29-year-old woman, the second youngest subject,


a 61-year-old man presented the second largest age dif-


ference (33 years)


Two 35-year-ol1d women married one 63- and


st. C .... ,.~~~ I A .-~~. Al 1 C -.- ,, --i-- r~t ~ *I a im n


91


one


,,


It,,,,,,,, r~











an 83-year-old man, took a 69-year-old wife.


Both of these pairs


were Negro.


When there


was an age difference


, It


was not always the groom


being older than the bri


8 of the 97 couples (roughly 20


percent)


as depicted in Table 6, the brides


were


older than the grooms.


McKain (1969:15) found that 15 percent of hi


brides


were


older than


their grooms.


In the


cases


of the 14- and 18-year spread, shown in


Table 6


Differences in A


of the Married Pair


n the


Cases


Where the Brides Were Older Than the Grooms


in Marriages


n Which at Least One of the Partners


Wias over Age 60, Performed in Alachua County,
Florida, 1966-1969, Inclusive


Years Number of
01der Cases__m____

1 2
2 3
3 2
4 2
5 3
6 1
7 1
8 2
14 1
18 1

Sum 18


Table 6, the former was a 47-year-old man married to a 61-year-old

woman and the latter was a 59-year-old man wedded to a 77-year-old





33





Unfortunately, the marriage license application forms in Florida


give only the locality of the bri


not any


and groom prior to the ceremony and


information on where the couple will be residing after the


marriage.
-J


The following lists were searched


n an effort to find the


current addresses of the subjects:


ty directories, ut


cities


billing lists, and telephone directories of the communities


n the


research area, the list of welfare clients of the Department of Social

Services for the counties involved, and the regional list of public

assistance recipients of the State of Florida Division of Family

Services.


The addresses of


some


of the subjects who lived farther away


were acquired by utilizing the telephone, dialing 1, then the area


code


, then 555-1212.


The information operator, upon request, always


supplied the street address where the telephone in that given party' s


name


was


located.


Many addresses were still not ascertained,


so a letter was


sent in care of the postmaster of each subject's last community of


residence with instructions to deliver it to that subject.


Each of


these letters had a return postcard enclosed in it requesting the


subject


to write down his present address and drop the card in a


mailbox.


If these cards


were


not returned, a personal visit was made


to every post office in the nine-county area to secure existing or












When this procedure failed to reveal the whereabouts of


all the


mining unknowns, inquiry


was made at various points such


as genera


stores and


service


stations in those towns in an attempt to get a


lead on those remaining


cases.


Information on the location of


some


subjects


was picked up in


each step


of the search.


In addition


some


postmasters and towns-


people were able to


say definitely that


some


of the subjects had moved


but for one reason or another had told no one where they had


gone.


Finally, a check


was made to


see


if any


of the subjects about


whom nothing was known


were


deceased.


This


was accomp


wished by inquiry


to the Bureau of Vital Statistics, Division of Health, in Jacksonvill

Florida.


Of the 97 legitimatized unions in the


area


and time span


indicated, the present whereabouts of the partners 60


years


old or


older is


as follows:


Number of


Cases


Ba lance


Tota
peop


number of marri


years


ages


invo


living


of age or older married


in Alachua County, Florida, during the
years 1966-1969, inclusive.


Subjects from communities


side the experiment


area


n Florida out-
who apparently


came to Alachua County for the wedding
and then returned to those communities
after the ceremony.

Subjects from states other than Florida
who apparently came to Alachua County for


re-












(continued)


Number of


Cases


Balance


Subjects from communi t ies inside the


experimental area who moved


Florida communities


to other


outside the nine-


county area after the wedding.


Subjec
inside
other


from Florida communities


the nine-county area who moved to


states


after the wedding ceremony.


Deceased subjects.


Subjects whose whereabouts remain
unknown despite exhaustive efforts.


Our experimental sample, therefore, consists of 77 subjects.


A tota


of 71 experimental subjects allowed the interviews to


be completed.


This represents a


92.2


percent response rate.


A tota


of 60


elderly neighbors were interviewed


representing a response rate of


95.2


as comparison subjects,


percent.


The investigation was carried out by personal interviews in

homes of the subjects. The field work was accomplished during the

months of June, July, and August, 1970.


Of the nine refusals, two


were


incomplete in terviews.


The re-


suits of these two were not tabulated with the 131 completed interviews.


One of the incomp


etes


resulted at the very end of the interview when


the subject suddenly reversed his desire to have the information re-

corded and demanded that the questionnaire be surrendered to him, he


immediately tore it up.


The interviewer outside later filled in a












questioning to continue that day or on any of the follow-up

ViSitS.


Generally speaking, however, the cooperation of the subjects


was very good


indeed.













CHAPTER THREE

FINDINGS


was stated in Chapter Two, 71 elderly newlyweds were chosen


the expe r imenn ta


samp


and 60 elderly persons


we re


selected


a comparison group.


All the subjects were interviewed in their own


homes and their responses to certain questions put to them are tabu-

lated and analyzed in this chapter.


Characteristics of the Research Samples


initia


is presented (Table 7).


y, the educational level of the subjects in both groups


This distribution demonstrates the range of


formal schooling among the respondents.


The data indicate


that the


experimental group had a higher percentage of subjects with no school-


ing, schooling to the fifth grade, and advanced educations, while


subj


ects


in the comparison group led in the classifications from a


sixth grade education to graduation from high school.


Information


was also collected on the primary lifetime occupa-


tlons of the subjects and these data are found in Table 8. The two

groups appear to be nearly evenly distributed on this variable, a]-


though the comparison group had more skilled craftsmen while the


expe rimenta


group had more lower level professionals.


The higher


prestige of the lower level professional


s is probably offset by the
















Table 7

The Educational Levels of the Experimental and Comparison Groups


Educat iona


Leve


Expe r i meant al


G


Group (N=71)
(Cum. %)


Comparison
roup (N=60)
(cune.


No Formal
Schooling


(8.5)


(5.0)


First Grade


Attended to Fifth
Grade Completed


29.6


(38.1)


21.7


(26.7)


Sixth Grade
Attended to
Eighth Grade
Completed


21.1


(59.2)


31.7


(58.',)


High School
Attended But Not


Completed


15.5


(74.7)


18.3


(76.7)


High School
Graduate


Some.


Tertiary


Ed ucation


15.4


(84.6)


(loo.0)


15.0


(91.7)


(100oo.0)


_ __ --- ---


_ -- ----









Table


The Primary Lifetime Occupations of the Experimental


and Comparison


Groups


Ex-peri mental


Primary Li fetime


Occupation


Group (tN=71)
) (Cum


Compari son
Group (N=60)
) (Cum. )


Housewi fe


30.0


(30.0)


Unskilled


Laborer


18.3


(46.5)


18.3


(48.3)


Service


Occupat ions


11.3


(57.8)


10.0


(58.3)


Semi skilled


\Worker


(62.0)


(61.6)


Clerica


Sales Categories


(69.0)


(66.6)


Si.le


Craftsman


(76.0)


18.3


(84.9)


Farm Owner or
Self-Employed


(84.5)


(91.6)


Lowe r


Profess ional


11.3


(95.8)


(96.5)


Upper
Professional


(100.0)


(99.9)


*Total cumulat i ve percentage


does


not equal 100.0 due to rounding.


Unskilled workers


= field and orchard workers, tenant farmers,


maids.


Service occupations


= restaurant and hotel employees, night watch-


men, railroad porters, military careerman, practical
nurses.


Semiskilled workers


factory


= gene


ral construction workers, truck drivers,


in;


~ssnnh~vI -


wnrker -


uuur


II '''"Y IIYI 1~Yl~t











ears


to be no great


socIoeconom


ic mrba


ance


between the two groups


which might introduce


a sampling b


las.


The data on the present


presented in Table 9.


employment statuses of both groups


are


There is little difference between the two


groups


on most of the


classic


icat


ons


but a disproportionate share of


Table 9


The Present Employmen t


Statuses


of the Experimental


and Comparison Groups


Present Employment Status


Experimental
Group (N=71)
) (cum.r)


Com~pa


r SOn


Group (N=60)
) (cum.


Housewife


28.2


30.0


(30.0)


Sti ll Employed Full-Time at
Primary Life Occupation


31.0


(59.2)


(36.7)


Retired from Primary Lifetime
Occupation but Still Working
Part-Time In It


(62.0)


(41 .7)


Retired from Primar


Lifetime


Occupation but Now Working
Full-Time in Another Occupation


(64 .8)


(43.4)


Retired from Primary Lifetime
Occupation but Now Working
Part-Time in Another Occupation


(64.8)


(46.7)


Fully Retired from
Work


Salaried


35.2


(10oo0.0)


53.3


C I co. )


the comparison group


was


experimental group subjects


fully retired and nearly five times


were


as many


still working full-time at their




41





whereas nearly hjlf of the comparison group subjects were not married.

It may also be true that marriage or remarriage may have been facill-


tated because the


subject had employment.


marital statues of


the subjects in both groups appear in Table


wa s


expected, most


Table


The Marital Statuses of the Experimental and Comparison Groups


Marital Status


Experimental
Group (N=71)
%) (Cum.r)


Compari


son,


Group (N=60)
) (Cum.


Married


85.9


(85.9)


60.0


(60.o)


Separated


37.5


Wi dowed


(60.0)

(96.7)


Divorced


(100oo.0)


(loo.o)


of the experimental group was still in a married


state.


Because


they


were originally selected on the basis of their being recently married,


should not be compared with the national figures


marital statues of old people in general.


as to the


However, the comparison


group


seems


to indicate that the aged people in the nine-county


area


of Florida roughly correspond to the marital statuses of other


aged


people in the United


States.


According to official estimates,


percent of al


are widowed (Ag


people over 65

ing, 1969:20).


years


Since th


age are married and 39 percent

e age floor in this study is


60. it is expected that sliahtlv more of the subjects would be


still












Regarding


sex,


41 subjects in the experimental group


were


and 30


were


females.


This


does


not indicate a reversal


in the aged


mortality trend but simply that many males in the


expert


mental group


were married to


women


were


too young to be subjects.


In the


compare son group,


slightly over 50 percent (31 subjects)


we re


females.


The figure for the whole United


States


years


old or


older is


7 percent female (Aging


l396S20) .


But it is


expected


that there would be a slightly h


percentage of males


in this study's comparison group, since the


floor for


still alive


section


owe r


by five


years.


Of the tota


131 experimental and comparison


group


subjects


74 or 56.5 percent


were


white and


or 43.5 percent


were


black.


the experimental group, 54 .9 percent


were


white while in


compare


ison


group


58.3


percent


were


wh i te.


The corresponding datum


for the State


of Florida


was 84.1


percent


white and for the nine-county


study


area,


which had 237,430 whites out


a total population of


was 77


.4 percent white (U.


Census,


1970:3-19).


This d i screpancy


can be partially accounted for by the fact that eight


the nine


refusals


intact


we re


by white persons; thus, if the samples had remained


, 59 percent of the subjects would have been white.


More of


this disparity is resolved by the fact that in Alachua County


Florida,


where most of the sample units were .drawn, only 69 percent of the


population over 60


years or aoe


was whi te (U.


-S


Census. 1960:ll-


wa s


,437, it











been 63. percent white if it had remained intact since all six


refusals in that group were by whites.


But all these


explanations


notwithstanding


of those


, it still remains true that a disproportionate share


people 60 years old or older who married in Alachua County,


Florida,


during the


yea rs


1966 to 1969, inclusive, were black,


although the difference is not great.


Testing of the Centra


Hypothesis


The major hypothesis of this work was that


a new marriage,


con-


traced late in life, will retard the disengagement process.


three measures of engagement used,


as outlined previously, were life-


space


measure f score, role count, and perceived lifespace


score


(see Appendix)


The decision was made that there must be a significant dif-


ference in all three measures to reject the null hypothesis.


each


A! though


of the tests, in itself, measures a dimension of disengagement,


they have been used in combination in past research; moreover, the


author desired that the testing be rigid as possible.


important variations are noted on a


Howeve r


single measure, they are d


* -n


discussed


at length.


A significant difference in the results of two out of


three measures is considered inconclusive.


The null hypothesis is


not rejected if a significant difference is noted in only one or none

of the measures.2





44





In the statistical analyses of all data, the lowest acceptable


level of confidence is the usua

are not considered significant


.05 level.


Any noted differences


at any lower confidence level


but any


clearly visible differences, albeit not statistically significant,


discussed.

The first of the three measures was the lifespace measure


frequency (or f)


score.


This was based on:


) the number of others


in the household (which


was multiplied by 30)


) the number of close


relatives, neighbors, and friends (which was multiplied by 30 if


seen daily


, by4 if


seen


weekly, etc.); 3) the number of coworkers


one talks to on the job, if the respondent is still employed (which


was multiplied by 20)


and 4) the number of other people


seen


week for certain


sum


became


spec


fic purposes (which was multiplied by 4)


the total lifespace f


score.


The results of this first test are presented in Table


, which


reflects a visible relationship in the direction predicted, although


the distributions failed to attain statistical significance.


close inspection, albeit the relationship


was


Upon


not statistically


significant, the data in the table reveal that nearly two-thirds (63

percent) of the comparison group respondents were concentrated in the

lowest two interactional columns while less than half (46 percent) of


the experimental group respondents were positioned there.


This


are


The













cC~5


0


CO O
- .*
c' O
O


a'~


. 9

0


9C 0D


a'



Cfl\D


U~O~


o..r






0O .,


cnr


00
O O
0 0
rnr"-


00
ifl',D
* -




ci
C-)


Q


h











clustering by the comparison group subjects in the lower columns


continued until the


scores


reached 300, with over three-fourths of


comparison


group


versus


less than two-thirds of the experimenta


group


account ted


for at that leave


\Wlith the majority


of the experi-


mental


group


scoring between 200 and 1,000 on the l i fespace


measure


and the majority of the comparison group scoring below 200, it


does


seem


to be true that the newly


ywed


aged subjects interacted


more


with


others in the house, relatives, friends


neighbors,


coworkers,


other specific people in general than did the other


aged
-s


subjects


in the stud


The second of three


measures


of engagement used in this study


was the total


role count.


The total role count


vias


based on:


1) number in household (scored


point if one other 1i


ves


zero


if respondent lives alone,


with him, and two points if two or


one


more


others live there)


2) number of close friends, relatives, and


neighbors (scored one point for any,


zero


for none); 3) specific


people (scored one point if any


are seen


zero


for none)


4) fellow


workers (scored


one


point


f respondent is employed


at al


none if


no longer working), 5) church (scored


one point if attends tw


month or oftener,


zero


attends less frequently); and 6) clubs


(scored


one


point if


a member of any,


zero


a member of none).


The sum


becomes


the role count.


The possible range is 0 to 9.


However


, no one in this study had


a zero


role count.










Tabl


e 12


The Total Role Counts of the Experimental
and Comparison Groups (Percentages)-
^^^^k~fTl^|~~tp^^h^^^l4^*fBt*B^ *4^H^|^HrU*'~l^^^*^lb~^**rhflB^J ^qikBq~ieW^lBB~^u~_^u~~rT^Ph-h-LlfH Jl^^^U~kIf W~rtq^^~~nnHUU^tW W ^ Mj ^q~^^l~f^WU^BU W 1T ^BWkJd4B^-lt~t --Uk^^i ^^ "- W 'll-- f- -U^ABr^Hi^^Bt^Hq^^^ ^~tt^J^'^i^^f^^ttf^ffW^~W^H~Hnfh^^^^f^k^~^^


To ta


Group


Role Count


Total


Exp.+


39.~4


22.5


(N=71) (1.4) (1.4) (5.6) (12.6) (28.1) (67.5) (90.0) (98.5) (100.0)


99.9
(99.9)


Comp.


28.3


16.7


20.0


00.0


(N=60) (0.0) (0.0) (6.7) (30.0) (58.3) (75.0) (95.0) (100.0) (100.0) (]00.0)


= 11.9


*Figur
+Total


in parent


percentages


heses d
do not


p > .05
enote cumulative
eaual exactly 1


percent
00.0 due


ges.
to rounding.


statistical significance, it does


appear


that the aged


newlyweds


had more


roles


as measured by number of close friends, relatives, neighbors,


friends, coworkers, household members, along with church and club activities


than had the other aged persons tested in this study.


ference can be noted in the first two


While little dif-


columns, the majority of the compari-


son group placed in the middle columns,


, 4, and


whe reas


the majority


of the experimental group fell in the highest columns,


7, 8, and


This


does


seem


to indicate that the newlywed aged subjects appear to have


more roles than the other aged subjects in this study.


observed


It can a


that 77.4 percent of the responses for the experimental group


fell in the 5-through-7


classes


as against only 65.0 percent for the


comparison group; and that while


58.3


percent


or wel


over half, of












group, were limited to that number of roles.


All this evidence points


in the direction of confirmation of the hypotheses,


even


though sta-


tistically the possibility of


samp


1i ng variabili ty


as the explanation


cannot be ruled out.


The third measure of engagement used in this study


was


the per-


ceived lifespace


score.


The perceived 1ifespace


score


was based on


the subject's evaluation of his international level at


in comparison with his present level.


age forty-five


The respondent scored more points


if more active now


some


points if activity


evel remained about the


same, and fewer points if he was more active in the past.


range was from 9 to 19 (


see


The possible


Appendix).


The results of the perceived


fespace test (Table 13) demon-


state a relationship in the predicted direction.


Although the two


groups


we re


fairly evenly distributed throughout the middle columns,


the comparison group had a disproportionate number in the lower two

columns, while the experimental group subjects had a larger share in


the two higher columns.


It does appear from the data that when the


respondents in both samples


were


asked to give their perception of the


difference in their activity levels now compared to these levels

age forty-five, the aged newlywed subjects perceived themselves


relatively more active than did the rest of the elderly respondents


at their present


ages,


though


again, the relationship did not achieve


S r -? .


ctz~tc rn c~ nfl r'nn'


- _-1


~t3t;ft;









Tabl


e 13


The Tota


Perceived Li fespace


Scores


of the Experimenta


and Comparison Groups (Percentages)*


Total Perceived Lifesoace


Group


Score


16-19


Total


Exp.
(N=71)(


22.5


22.5


) (7


(81.7)


9.9
(91.6)


(95.8)


(100.0)


100.0
(100.0)


Comp


21.7


11.7


00.0


N=60)(23


(46.6


4.9)


(96.6)


(100.0)


(100.0)


X2 = 1.7 df = 7 p .05
. *Figures in parentheses denote cumulative percentages.


three measure


of disengagement


, the null hypothesis cannot be rejected.


However, there


was a tendency in the data in all three tests toward


confirmation of the central hypothesis.


Testing of the Subhypotheses


Subhypothesis A stipulated that married subjects in the compari-


son group


would be more engaged than unmarried subjects in that group.


responses


of these subjects


were


tabulated and analyzed in


same


manner


as was


case


with the central hypothesis.


For this


analysis,


the widowed, divorced, and separated columns shown in Table 10


were


collapsed into


a single unmarried


category.


For the comparison


group, then, this means 36 respondents (60?%)


are


classified


as married


and 24 respondents (40%)


as unmarried subject


An ,*nc r rn eiat t -I, a n:o Fcno


erR~~


1.17F


F


a











Table 14


The Lifespace Measure f


Scores


of the Married and Unmarried


Subjects in the Comparison Groups (Percentages)*


Tota


Score


on Lifespace


Measure


Group


Tota


0-99


100-139


200-299


300-399


400o-419 9


over


Married


(N=36)


22.2


(22.2)


(58.3)


16.7
(75.0)


(77.8)


(89.9)


(100.0)


100.0


(100.0)


married


(t=24)


41.7
(41.7)


29.9
(71.6)


12.6


(84.2)


(92.6)


(96.8)


(loo100.0)


00.0


(loo100.0)


2.2


*Figures in parentheses deno


cumulative percentages.


relationship in the direction predicted.


Nearly twice the percentage


or unmarried subjects


as married


ones


scored


in the very lowest life-


space


category while the incidence of married subjects in the scoring


columns of 200 and above


was nearly double that of unmarried subjects.


Hence, it


does


appear


that the married subjects in the comparison


group interact


slightly more with other people in general than do the


unmarried subjects in that group, but again the relationship was not

strong enough to be significant statistically.


The role count results presented in Table


5 reveal


a tendency


for the data on this measure to offer support for the hypothesis.


majority of the unmarried subjects fell


in the lower three columns and


the mainritv of the married P nlh pn tfha Inner threP* thus. t he r]dt-










Table


The Tota


ole Counts of the Married and Unmarried Subjects
n the Comparison Group (Percentages)*


Marital
Status


Total Ro


Count


Total


Married
(N=36)


(8.3)


(16.6)


(44.4)


16.7
(61 .1)


30.6
(91.7)


(1oo0.0)


100.0
(100.0)


married+
(1=24)


16.8


(4.2)


(49.4)


(m.3)


(96.1)


(100.3)


(100.3)


100.3
(100.3)


X2 = 7.7 df = 5 p > .05
*Figures in parentheses denote cumulative percentages.
+Total percentages do not equal exactly 100.0 due to rounding .


specific acquaintances, re


atives


, neighbors, coworkers, others in the


house, and are more active in clubs and church than the unmarried


subjects in the comparison group


even


though the distributions


were


not sufficiently different to be significant at the acceptable level

of confidence.

None of the comparison group scored higher than 16 on the per-


ceived lifespace measure (Table 16).


The distributions show that the


chances of an unmarried subject being in the lowest scored column were


more than 50 percent greater than for


a married subject and the inci-


dence of married subj


ects


in the uppermost category


was considerably


higher than that of the unmarried subjects.


Hence,


elderly married


rn~nla











Table 16


Tota


Perceived Lifespace


Scores


of the Married and Unmarried


Subjects


in the Comparison Group (Percentages)*


Mari


Tota


Perceived Lifespace


Score


-Total


Status


Married
(N=36)


(19.4)


(44.4)


27.8
(72.2)


(86.1)


(94.4)


(loo.0)


100.0
(100.0)


married+
(N=24)


29.9
(29.9)


20.8
(50.7)


12.6


16.8


(80.1)


(84.3) (100oo.0)


(100.0)


= 6.9


*Figures in parent


heses


p .05


denote


cumulati


percentages.


+Total percentages do not equa


exactly 100.0 due


to rounding.


difference


fidence


was


insufficient to be significant at the acceptable con-


evel, however.


Since


a difference in marita


status in the comparison group


was not shown to be significant in any of the three


measures


of dis-


engagement, the null hypothesis cannot be rejected.


Again, however,


the data demonstrated a tendency in all three tests toward confirm-

ing the hypothesis.

Subhypothesis B maintained that female subjects in the experi-

mental group will be more engaged than male subjects in that group.


The lifespace measure results


are presented in Table 17.











Table 17


The Lifespace Measure f


Scores


of the Male and Female


Subjects in the Experimental Group (Percentages) '


-~~~ ----~


Total f


Score


on Lifespace Measure


0-39


200-


300-


400-


Total


500-


1000-


2000


1999


Over


Male+-
(N=41)


17.1


22.0


22.0


(17.1) (39.1) (61 .1) (68.4) (73.


19.5 4.9 2.4 100
(92.8) (97.7) (100.0) (100


Female
(N=30)


20.0


36.7


16.7


0 13.3 3.3 6.7 3.3 100.0
4) (86.7) (90.0) (96.7) (oo100.0) (oo100.0)


- 2 1


*Figure


s in par


entheses


denote


cumulative


percentages.


+Total percent


ages


do not


ezua


exactly 100.0


due to rounding.


The role count by


sex is depicted in Table 18.


The differences


between the distributions


were


not statistically significant.


The per-


centage
-d


of ma


was


greater in the lower three role count columns but


Table 18


The Total Rol


Counts of the Mal


and Female


ubjccts


in the Experimental Group (Percentages)


Tota


Role Count


Total


i-a le+


34.1


(2.4)


26.8


12.2


994. 9


4) (7.3) (14.6) (24.4) (58.5) (85.3) (97.5) (100.0)(100.0)


(20. 0) (56.7) (73 it)












was also greater in the three higher role count columns.


centage of females


The per-


was greater in the middle three role count columns.


The weight of the


extremes


cancels out the weight


of the mean, and


the data remain inconclusive.


The perceived lifespace


score


sex is presented in Table 19.


Table 19


The Total


Fema


Subj


Perce
ects


ived L


fespace


Scores


in the Experimental Grou


of the Male and


(Percentages)*


Total Perceived Lifespace


Score


16-19


Total


Male+
(N=41)


14.6


22.0


26.8


19.5


99.9


(14.5) (36.6) (63.4) (82.9) (85.3) (90.2) (95.1) (100.0) (100.0)


Fema le
(N=30)


10.0


16.7


10.0


16.7


16.7


(10.0) (33.3) (50.0) (60.0) (76.7) (93.4)


100.0


7) (100.0) (100.0)


=3.6


p > .05


*Figures


in parent


heses


denote


cumulative


percentages.


+Total percentage does not equal exactly 100.0 due to rounding.


As was the


case


with the two previous tabulations, the variances


between distributions


were


not significant statistically and the data


in the table, when studied, also appear neither to confirm nor to deny


this inconclus


iveness.


A difference i n


sex


did not prove to be a significant variable


-. 4r 4i*r










Testing of the Workina Hypotheses and Corollaries


The first working hypothesis stipulated that the greater the

number of living children the aged subjects have in either group, the


less will be their disengagement.


The initial corollary stipulated


that,if the null hypothesis can be rejected, the relationship will be


stronger in the experimental group than in the comparison group.


breakdown of the number of living children in both groups is presented


in Table 20.


As can be


seen,


more of the experimental group had no


Table 20
Number of Living Children in the Experimental
and Comparison Groups (Percentages)>'


Group


Number of Livinq Children


Total


9 or
More


Exp.+
(N=71) (


.4) (43.7) (71.9) (77.5) (84.5) (


8 1.4 2.5 2.a 5.6 99.9
3)(88.9)(91.5)(94.3) (99.9) (99.9)


Comp.
(N=60)


15.0 33
!5.0)(48


13.3


00.0


)(76.6)(89.9) (93.2) (94.9) (94.9) (96.6) (98.3)(100.0)(100o.0)


*Figures in parentheses denote cumulative percentages.


+Total percentages do not eaual


exact


v 100.0 due to rounding.


children.


The experiment


group also had


a greater incidence of four


or more children with 22.4 percent distributed in those columns with


percent for the comparison group.


The last five columns are collapsed into a single "5 or more"












0O 0
a


NO


0 0


-C'
a a
p-a'
-0~'~


C-'


oco
a
ow


oc


C-.
*-co
a a
-03
-C)


OCNJ


cFC1~J


00
a
LflO
0~'


r-rn
* a








* C

'0





C-
(Nj a.~


C-


0 0


0 0








O r-
o0o


0 0


0 0


00
*
00
0



c-'0



a

0


0 0


OO
0 0

0 0
N-


tn-t
0~




C-'
En C'J
C
r~cO
c'4oZ~


00


00
* e


00
* C


C
0
-4
Lfl

0~'-~
E
0
0


~F2

















0 0


0 0


0 0


00


0 0


LAO~


0 L(







o o
** *
0 0






o o
*
CM r



00o


O 0


C-
au-I
*
oc"J
kO




C'
Otr~
* 0
ON
'-0





C-.


0 0
* L
LPNrN
C OJ


0 0


O *


O 0
00


00
* S
00


0 0
S


C-.
CO


00


00


0O 0



* O
0 0
0






0 0


O 0


OIO
0 0

Or\ CTA
c^ en


OO
00
3~0^


0 0


0 0


C'
00
S

C~\~0





C'
00


0 0
* S
C\C\\


00


0 0
0 0
0 0
uL/


0 0


C-'
00
0 0
u-'u'
CNJCNI


-o


LA
0
*

A

taO


A


N
0*
Ho-

C)

(-5


Utnc
C)
NJ tflO
>c 0)-u

4J(fl
.0cc)


0044
* Q~c
C!) C)
Cu
U.-.-
o CI
(/) u, a.
S-c)
LL-
cont

~ a
OIL 1--.
o.-~ +~














o'a'~


-R-


0


o0


C-'
-0\

-a'
r-o~
4%-


C-
4Th-

Lr~4


-c


r-. -



C-.
u~Lr\


Ss


- -D


e -













0 0


O


0 0


0 0


0 00

01 O


00


0 0
** *
0 0
0





urnO


0 0
0




0 0
* S
0 0
0






0 0


00


0 0




0 0
**
0 0





0 0


00


0 0
*
0 0
0





0 0
* a


-S


0 0
* 1
0OLCI


U\0
* S

kOh~




OLC\
* S
ON




--S
tnt"
* S
NC'J
- -



C-
00
* *
00


00
* ft
00


0 0
* *
0 0
CO




00


OO
00

00





00
* 6
00


00
* S
00





00
O o

*



00
OO



OO
00
OO*
00o


0 0
* a
0 0
0





S
0 0
0






0 0


0 0

0 0

0 0


0


00
O*
00
OO-




00
OO*
00




00
* S

00
OO-.


*
o r
cm





00
* a
oO







o o
* *






0 o


0 0


00


00
* S
00





00
* S
00




00
* S
00


00
0
I-





S
0 0
0
OO

Oo


C-'
00
S S
oLd-'


00


C--
00
S S
LflIS\
c'JN



C-'
00


00


O O


-ZrQ


S. ,-
toeD

DC
00.)
LE
0*-
L
SZC)
4-sct
ox
cow





































~0-
v-tn


00
4-,

cc-I
a-u

--C)
-in-
4-
4-
ocn
a


.flL


S S


0 0
0









-

0


C-
'-Dc-'
S

co


C-


-N


C-,
a~u~
0 5


ocr


00*


ocr~
* a

cm







a


4 Sn


-r u1


tr-


0O


0 0
O *
0 0
0


00S


0


r


S0


tO
0








0 0


C-


A 0
.-0
-0


0~r\


* 4








C- \
* S
F .> -
W1 r^


- 0


C-
0 0
*

LC\O






0 0
* a


0 0


- A L


0 0

in lLa


0
* S


C-
00
S 4
LaO
-a,


00
a a
Lt\ Lil



LO
v0







0 0


0o-


-03







C-
'flit'


O-
*a


- -


00
1 1
u\u


In LA


C-
0)0
EC'J
a-fl


a










0


0o S


O


00


00


0 r


0

0
rO






00c


C-
00
*
Cu-'







C-
00
* S
flu'
CNJr-%


0


oo


C-
tf'0


O Lf
* 9





Oth-
00








S


C-
00
C
Ott'







C-
LAO


C-
LA LA


0 0


0


00
O *
0 0






CO
0 0
* S
00
\^


0 0
* S
0 0








0 0


0 0
* S


0


- -
C-
0 0


00
*OO
0 0
0







00
o S
0 0
0






0 0'


0 0


00
S S
00


OO
00
S S
00


00
* S






o'c



00


0 0

O o
cn



o o


0 0


0 0
* a
en o,


0 0
* *


C r

mm


00


OO
00

0 0
LA


O-
00
*
00




OO







OO
0 0
* a
0 0
C.o.







00


II
C
(-'C

S.
-
0
0
S

V

C


II
LI)
00
S.
- -en
Ji.- 4-)
~0c
*0C)

1~~
10
a
aa
C)

a -
44
no rtj
"-' -S
II
II E


a a
U)

:J C.
00-
>CLJ
C-Jo
U
C S I-













The distributions in Tables 21 and


were


not significantly


different and the data in them, on being scrutinized, appear to reflect


the absence of the predicted relationship.


All the chi


squares


pro-


duced from Table


were significant, but


since


a difference in


number of living children


wa S


shown to be significant on only


out of three measures of disengagement, the null hypothesis cannot


rejected.


By definition, this also makes the corollary untenable.


However, when


a C test


was performed on the chi squares from the


measure which produced a statistically significant difference, it


was shown that the relationship


was stronger for the comparison


group


than the experimental group, which


s in the direction opposite from


the prediction.


Therefore


, though the null hypothesis cannot


rejected, the number of living children did prove to be


a significant


variable on one measure of disengagement (perceived lifespace)


there


exists


a partial tendency toward the confirmation of the hypothe-


S'S.


Surprisingly, however, though the data tend toward affirmation


of the hypothesis in general


, they a


seem to refute the corollary.


Hence, it appears, at least in part, that children interfere with


disengagement process of all aged people but more


who are not in a new marriage situation.


so with older people


This particular study did


not gather any explanatory data on this situation but it may be


a 1 -. .1 L. 3 .4 ... I L. L


S a- -


. ~ ~ L r.l -z ~z --.- -- .-A


one


one


_^ _










perhaps, merely for estate purposes; whereas an old person without

close family ties may tend more to marry again simply for companion-

ship due to his loneliness.

The second working hypothesis stated that the greater the


number,


living siblings an aged person has in both the experimental


and comparison group


, the less will be his disengagement.


corollary stipulated that if a significant difference is noted, the


relationship will be stronger in the


compare son.


experimental group than in the


Initially, it will be of interest to tabulate the number


of living siblings for both groups.


This is accomplished in Table 24.


Table 24

The Nlumber of Living Siblings in the Experimental


and Comparison Groups (Per


IlLt


ages)*


IIII-. -- I-1I .I


Group


Number of Living Siblings


Total


9 or
more


Exp.+-
(N=71) (


23.9
23.9


22.5 18.3 9.9 5.6 7.0 5.6 0.0 2.8 4.2 99.8
)(46.4)(64.7)(74.5)(80.2)(87.2)(92.8)(92 .8)(95.6)(99.8)(99.8)


Comp.+
(N=60)


28.3
(28.3


15.0 15.0 8.3 13.3 8
) (43.3) (58.3)(66.6)(79.9) (88


3.3 1.7 3
) (91.5) (93.2) (96


3
)(99


99.8
)(99.8)


*Figures in parentheses denote cumulati


percentages.


+Total percentages do not equal exactly 100.0 due


to rounding.


The data indicate that there


was very little variation in the number










three tables.


These distributions, then, on the three measures of


engagement by the number of living siblings,


are presented in Tables


and 27.


Table


Total Lifespace Measure f tScore by Number of Living Siblings for
the Experimental and Compoarison Groups Combined (Percentages)*


Number


Living
Siblings


Total L


0-99


100-
199


200-
299


ifespace


300-
399


IMeasu re


400-
499


500-
999


1000-
1999


Total
2000 or
more


0
(N=34)


32.4 35.3 17
(32.4) (67.7) (85


2.9 2.9 2.9 2.0 2.9 100.0
) (88.4) (91.3) (94.2) (97.1 ) (100.0) (100.0)


1
(N=25)

2
(N=22)


24.0 44.0 4.0 4.0 8.0 8.0 8.0 0.0 100.0
(24.0) (68.0) (72.0) (76.0) (84.0) (92.0) (100.0) (100.0) (100.0)

27.3 22.7 27.3 4.5 18.2 0.0 0.0 0.0 100.0
(27.3) (50.0) (77.3) (81.8)(100.0) (100.0) (100.0) (100O.0) (100.0)


3
(N=12)


0.0 25.0 16.7 0.0 8.3
(0.0) (25.0) (41.7) (41.7) (50.0)


.0 16.7 8.3 100.0
.0) (91.7) (100.0) (100.0)


4+
(N=12)


33.3 25.0 25.0 0.0 0.0 0.0 8.3
(33.3) (58.3) (83.3) (83.3) (83.3) (83.3) (91.6)


8.3
(99.9)


99.9
(99.9)


5
(N= 10)

6
(N=6)


or more
(N=10)


20.0 20.0 30.0 10.0 0.0 20.0 0.0 0.0 100.0
(20.0) (40.0) (70.0) (80.0) (80.0)(l00.0)(0O0.0) (100.0) (100.0)

33.3 16.7 16.7 33.3 0.0 0.0 0.0 0.0 100.0
(33.3) (50.6) (66.7)(I00.)(O0.0) (100.0)(l00 O.0) (100.0) (o100.0)


0.0 30.0 10.0 0.0 30.0 20.0 0.0 10.0 100.0
(0.0) (30.0) (40.0) (40.0) (70.0) (90.0) (90.0) (100.0) (100.0)


X2 = 107.3 df = 49 D <.01. C = .63












The contingency coefficient of .63, which was calculated for


Table


, is indicative of a moderately strong relationship.


Although


the tabulation summarizes the data for both groups combined, cht


squares


we r e


computed from the raw data on each group separately;


for the experimental group it was significant at the .001 level, while


for the comparison group alone it was not significant.


From this it


appears, in this first test at least, that the data tend to demon-


state a relationship


and that it


s in the direction predicted.


The results of the second-test on the importance of the number

of living siblings to the total role count are presented in Table 26.

Again, although Table 26 only summarizes the data for both groups,

chi squares were computed on the raw data for each group alone;


however, neither chi square


was of sufficient magnitude to be signifi-


cant at the .05 level of confidence.


The outcome of the perceived lifespace


giving siblings is depicted in Table


score


As in Tables


by the number of


and 26,


distributions in Table 27 represent the total for both the experimental


and comparison groups taken together; however, the ch


i square statistic


was again computed on the raw data from each group by itself and in


both


cases


the chi square produced was not significant at the accepta-


ble confidence level.

Since a difference in the number of living siblings was shown

*njr kcs c" nn rtirn l C-i -i.^- k,-^\V k ,,V r-rn/vlr~c- p. ant 1^^ ,-n an 4^v -. r,'F *f-l rao^^ mot^ *^C irac /











Table


The Total Role Count by
the Experinm ntal and Com


the
pa r I s


Number of
on Groups


Living Siblings for Both
Co:bined (Percentages)*


- -- -- --.-~- -.- -~


Total Role Count


Living
Siblings


Total


1 __2 3 4 5 6 7


0+;


2.9 0.0
(2.9) (2.9)


5.9
(8.8)


23.5
(32.3)


26.5
82.3)


14.7 0.0 2.9 S9.9
(97.0) (37.0) (99.9) (99.9)


1
(N=25)


0.0 0.0
(0.0) (0.0)


4.0 20.0
(4.0)(24.0)


28.0
(80.0)


16.0 4.0 0.0 100.0
(96.0)(oo100.0) (oo100.0)(1oo00.0)


2
(N=22)


0.0 0.0
(0.0) (0.0)


9.1 9.1 13.6 40.9 22.7 4.5 0.0 100.0
(9.1)(18.2)(31 .8)(72.7) (33.,) (00.0)(100.0)(100.0)


3+


0.0 0.0
(0.0) (0.0)


8.3 0.0 0.0
(8.3) (8.3) (8.3)


0 33.3
3) (66.6)


33.3
(S9.S)


0.0 99
(99.9) (S9


(N=12)


0.0 0.0
(0.0) (0.0)


0.0
(0.0)


.0 16.7
.o) (4m.7)


.0 8.3 0.0 100.0
.7) ( 100.o) 0Ioo.)(10oo. o)


5
(N=10)


0.0 0.0 10.0 10.0 20.0 30.0 20.0 10.0 0.0 100.0
(0.0) (0.0) (10.0)(20.0)(40.0)(70.0) (90.0)(00.0)(3)00.0)(O0O. )


6
(N=6)


0.0 0.0
(0.0) (0.0)


0.0 0.0 50.0 0.0 50.0 0.0. 0.0 100.0
(0.0) (0.0) (50.0) (50.0)(100.0)(100.0) (100.0) (100.0)


or more 0.0 0.0
(N=10o) (0.0) (0.0)


0.0 0.0 30.0 40.0 20.0 10.0 0.0 100.0
(0.0) (0.0) (30.0)(70.0) (90.0)(00.0)(100.0)(00O.0)


X2 = 45.9 df = 48 p > .05
*Figures in parentheses denote cumulative percentages.
+Total percentages do not equal exactly 100.0 due to rounding.










Table 27


The Total Perceived Lifespace Score by the Number
Siblings for Both the Experimental and
Comparison Groups Combined (Percentages)


Number


Total Perceived Lifespace


of Living


Score


Living
Siblings


0
(N=34)


Total


16-19


9


1 )4.7 26.5 29.) 14.7 5.9 5.9 2.9 0.0 100.0
(14.7) (h1 .2) (70.6) (85.3) (91 .2) (97.1) (100.0) (100.0) (100.0)


1
* (N=25)


20.0 28.0 12.0 12.0 8.0 12.0
(20.0) (48.0) (60.0) (72.0) (80.0) (92.0)


0.0 8.0 100.0
(92.0) (100.0) (100.0)


2+
(N=22)


7 27.3 27.3 13.6 4.55
7) (50.0) (77.3) (90.9) (95.4) (99.9)


0.0
(99.9)


0.0
(99.9)


99.9
(99.9)


3+
(N=12)


16.7 8.3 25.0
(16.7) (25.0) (50.0)


8.3 8.3
(91.6) (99.9)


0.0
(99.9)


0.0
(99.9)


99.9
(99.9)


4+
(N=12)


8.3 33.3 16.7 16.7 8.3 8.3
(8.3) (41.6) (58.3) (75.0) (83.3) (91.6)


0.0
(91 .6)


99.9
(99.9)


5
(N=o10)

6
(N=6)


20.0 10.0 20.0 10.0 0.0 30.0 10.0 0.0 100.0
(20.0) (30.0) (50.0) (60.0) (60.0) (90.0) (100.0) (100.0) (100.0)

0.0 16.7 33.3 16.7 0.0 33.3 0.0 0.0 100.0
(0.0) (16.7) (50.0) (66.7) (66.7)000.0) (100.0) (100.0) (100.0)


7
or more
(N=10)


30.0 10.0 10.0 0.0 10.0 10.0
(30O.0) (40.0) (50.0) (50.0) (60.0) (70.0)


20.0 10.0 100.0
(90.0) (oo100.0) (100.0)


X2 = 68.2 df = 49 p .05
*Figures in parentheses depict cumulative percentages.
+Total percentages do not equal exactly 100.0 due to rounding.





68





the groups taken separately), the null hypothesis cannot be rejected.


By definition this


makes the corollary untenable.


However, the


results of one test


1 ifes


pace


measure f score) were in keeping with


the h


ypot


hesis and the relationship was certainly stronger in the


experimental group; thus, it can be said,


as well


that the data


tended to confirm the corollary.


In human terms this means that old


people with a greater number of brothers and sisters


see and talk to


more people generally than old people with fewer brothers and sisters.


It a


means that aged newlyweds with the


same


number of brothers


and sisters


as aged persons who are not newly wed tend to


see and talk


more to other people.

The third working hypothesis stated that urban aged subjects


will be more engaged than rura


aged subjects.


The corollary stipu-


lated that if a


gnificant difference is noted, the relationship


will be stronger in the experimental than in the comparison group.


The di


stributions for the lifespace measure f


score


by resi-


dence (Tab


28) did not vary enough to achieve statistical significance.


The data, on examination, do not seem to revea


ship, hence the lack of statistics


any degree of relation-


significance appears to be corro-


borated.


The results


of the role count device by rural/urban residence


are exhibited in Table 29.


Even though the ch


square statistic


generated from the distributions was not significant, the rural re-




69





Table 28

Total Lifespace Mi asure f Score by Rural and Urban
Residence for Both the Experimental and Comparison Groups Combined
(Percentages)^'


* Resi-
dence


Tota l Lifespace


0-99


100-


300-


% ;?r~5L I-e


f Score


500-


1000-


2000- 3000


Total


199 299 39490013 9


over


21.2


(N=66) (21.2) (54.5)


16.7 6.1 7.6 10.6 4.5 0.0 0.0 0.0
71.2) (77.3) (84.9) (5.5) (100.0) (100.0) (100.0) (oo00.0)


Rura 1--


26.2


27.77


"18.5


(r;=65) (26.2) (53.9) (72.4) (75.5) (84.7) (89.3) (93


X- : 0.3 df = 8
*Figures in parent


eses


ict cumulative


percentages.


+Total


percenteqes


do not


equaE exactly 100.0 due to rounding.


Table 29

The Total Role Count by Rural and Urban Residence
for Both the Experimental and Comparison Groups Combined
(Percentages)o

.1- -- -.mim li-II*


Total Ro


Count


__ _--,, ,Tota1t


Urban+-
(N=66)


(0.0) (0.0)


0.0 16.7 18.2 27.3 25.8 10.6 1.5 100.1
(0.0) (16.7) (34.9) (62.2) (88.0) (98.6) (100.0)(100.1)


Rural
(N=65)


10.8


(1.5) (


12.3


24.6


30.8


16.9


100.0


Urban


1.5 4.6 100.5
(95.9)(100.5)(100.5)


ResI -
dence


.5) (12.3) (24.3) (49.2) (80.0) (96.9)(100.0) (100.O)(100.0)


i, nn,











the urban respondents led the rural ones in the three higher role


columns.


Thrlis


seems


to give support to the hypothesis and in the pre-


dicted direction.


The chi square cormputed from the third test involving the


per-


ceived lifesDace


score


by rural or uruan residence of the subjects


was zot


significant, and the data within the distributions also appear not to

demonstrate the predicted pattern (Table 30).


Table 30


Total Perceived Lifespace


Sc re


by Rural and Urban Residence


for the Experirnmntal and Comparison Groups Combined
(Percentages) ,

*T r ri! r Iil ir ~i I~ i~~r ** 'ii ~~iiin ii ii 1 l ll 1*;.- -i_-_|- .illi.L 11.*** *- ,, -- n ^ -_r "il- ... i |1" 11_ ~'" ._41 ** ~lril__ i ^^_ "r '"~ U'.-L 11^ *r L_ T -" _1 -*** .l *i-ri -im U- -"" ir.n -- -- --- ---^- _-. j -- .__ ._ ^ _-.i __ -_ ^ _


Total Perceived Lifespace Score


Res i-
dence


17-19


Total


Urban


19.7


21.2


19.7


100.0


(N=66) (19.7) (40.9) (60.6) (75.8) (81.9) (91.0 ) (94.0 ) (58.5)(100.0)(100 .0)

Rural 15.4 24.6 24.6 13.8 6.2 12.3 3.1 0.0. 0.0 100.0
(N=65) (15.4) (40.0) (64.6) (78.4) (84.6) (96.9)(100.0)(loo.0)(loo0.0)(100.0)


= 0.45


p> .05


Figures in parentheses depict cumulative percentages.


Since an urban-rural difference was not statistically demonstrated

to be significant for any of the three measures of engagement, the null


hypothesis cannot be rejected.


By definition, this also makes the corol-


lary untenable.


However, on one measure (role count) there


seemed


to be


COun t











suggests


that aged persons living in cities do tend to have more con-


tacts with other people on


a regular basis than elderly people living in


the country.

The fourth working hypothesis stated that the subjects in both


groups


~.ave


spouses


under the


of 60 will be more


engaged


than the


subjects


whose


spouses


are 60


years


or older.


The corollary stipu-


lated that if


a significant difference is noted, the relationship will be


stronger in the experimental than in the comparison group.


The outcome of the lifespace


measure


score


the age of


spouse


for the


subjects


of both groups combined


is displayed in Table 31.


Though


Table 31


Total Lifespace fea


sure


by Age of


Spouse


for Both


the Experimental and Comparison Grou


Combined (Pe


rcen


tages)*


Total Lifespace


0-99


100-


200-


4,


300-


Score
500-


1000-


2000


Total


1999


over


60 or Over+
(N=62)


22.6 37.1 17.7 3.2
(22.6) (59.7) (77.4) (80.6) (


4.8 6.5 4.8
85.4) (91.9) (96.7)


3.2
(99.9)


99.9
(99.9)


Under 60
(N=34)


17.6 14.7 23.5 5.9 11.8 14.7 5.9 5.8 100.0
(17.6) (32.3) (55.8) (61.7) (73.5) (88.2) (91.4) (100.0) (100.0)


= 6.6


*Fig


ures


in parenthes


p> .05
represent


cumulative


percentages.


+Total pe


rcengages


do not equal exactly 100.0 due to rounding.


Ij I *I1 II'. 111 ill''' If1'tL l 7~ar


n r f n nnr t-Hff n nPrent 0nnsn s- t;rl n F/ crrzf~1 cnnfb-,ri


r i~un


400-











wi th


spouses


over 60


years


old falling into the two lower columns on the


table and the subj


ects


with younger spouses leading on all the higher


columns, the configuration of data


seems


to lend support to the hypothe-


SIS.


Simplistically, this means that the younger the


spouse


an aged


person has


the more people he will know.


Table


presents the results of the ro


count


test


Table


Total Role Count b
Comparison Groups


of Spouse


Sepa


for the Experimental and


rately and Combined


(Percentages


Age of
SpouseGroup


Total Role Count


Total


60 or Exp. 0.0 0.0 5.7 14.3 14
-Older (N=35) (0.0)(0.0) (5.7)(20.0)(34


48.6 17.1 0.0 0.0 0.0
) (82.9) (100.0) (100.0)(100.0) (100.0)


Comp.+


25.9


18.5


Both+


11.3


19.4


100.1


(N=62) (0.0)(0.0) (8.1)(]9.4)(38.8)(74.3) (96.9)(100.1)(100.1)(100.1)


Under Exp.


21.4


35.7


21.4


00.0


(N=28) (3.6)(3.6) (3.6) (3.6) (17.9) (39.3) (75.0) (96.4)(100.0)(100.0)


Comp.+
(N=6)


16.7


16.7


50.0


16.7


00.1


(0.0)(0.0) (0.0)(16.7)(16.7)(33.4) (83.4)(100.1)(100.1)(100.1)


Both+


20.6


20.6


99.9


(N=34) (2.9)(2.9) (2.9) (5.8)(17.6)(38.2) (76.4) (97.0) (99.9) (99.9)


Both Groups:


11.44


=8


p> .05


- --2 --- a


(N =2 7), (0 0) (0 O ) ( 1 1 1 ) ( 1 8 5) (44 4 ) (62 3 ) (9 2 5 ) (9 9 9) (99 3 )












spouse.


The situation introduced here is unusual in that the distribu-


tions


for both g


groups


combined proved not to vary sufficiently to be


statistical 1


gnificant, which tends to refute the hypothesis, and the


data in the


taken


C -\


do not

the chi


seem


to contradict this finding.


square


for the experimental group


However, when

was statisti-


cally significant


while it


lwas not for the comparison group.


This


be interpreted


as supportive


of the corollary, which


means


that


aged newly


yweds


with young


spouses


tend to be more


active


in church


clubs and have


more


friends,


acqua intances,


neighbors


, relatives,


workers, and others in the house than do aged people in general who also


have young


spouses.


The distributions


resulting from the perceived lifespace


score


measure


spouse


appear


in Table


Notwithstanding that the


Table


The Total
the Exper


Perceived
mental and


Li fespace
Compare so


Score
n Grou


Age


of Spouse for Both


Combined


(Percenta


ges) w


Age of


Spouse


Total Perceived Lifesoace


Total


15-19


60 or Over


= 62)


21.0


25.8


21.0


(21.0) (46.8) (67.8) (83.9) (


.7) (95.2) (100.0)


00.0


(100oo.0)


Under 60
(N = 34)


23.5


11.8


11.8


(5.9) (29.4) (58.8) (70.6) (79.4) (91.2) (100.0)


00.0


(loo.o)


Cc-


- -- - --





74





variance between the distributions did not reach statistical significance,

a greater percentage of subjects with older spouses occupied the two

lower columns of the table while a greater percentage of subjects with


young


spouses


occupied the three higher columns.


This


means


that


aged


people in


general


who have young


spouses


perceive themselves


as relatively


more


active now when compared with their younger selves than


do other aged


individuals whose


spouses


are over


years


aY a


Since a difference in the


spouse


was not shown to


'statistically significant for both groups in any of the three


engagement, the null hypothesis cannot be rejected.


measures


By definition, this


makes the corollary untenable.


However, on two of the


measures


(lifespace f


score


and perceived lifespace score) the data, after


inspec-


tion, appeared to confirm the hypothesis to a degree.


Also in terms


the role count test, the data lent credence to the corollary.

The fifth and final working hypothesis stipulated that in both


groups


whites will be more engaged than blacks.


The corollary


stated


that if


a significant racial difference is noted, the relationship


will be stronger in the experimental than the comparison group.

The distributions of both groups on the lifespace measure


score


race


did not differ enough to achieve statistical significance


and the data within those distributions, when studied, do not appear


demonstrably vary (Tabl


The r










Table 34

The Lifespnce Measure f Score by Race for All Subjects
(Percentages) ''


~-- ~r -- --~~rt--r*- --cx 1C---.- - - s- -v -


Total Lifespace f Score


0-2


100-


200-


300-


too 3-


500-


1000-


2000- 3000


To t a


! 399 499 999 1999 2999 over


White+
(N=74)

Black+
(N=57)


23.0 27.0 16.2 2.7 8.1 10.8 6.8 1.4 4.1 100.1
(23.0) (50O.0) (66.2) (68.9) (77.0) (87.8) (94.6) (96.0)(100.0)(100l.1)

24.6 35.1 19.3 7.0 8.8 3.5 1.8 0.0 0.0 100.1
(24.6) (59.7) (79.0) (86.0) (9..8) (98.3)(100.1)(100.1)(100.1)(100.1)


= 4.2


p> .05


-Parenthetical figures represent cumulative percentages.
+Total percentages do not equal exactly 100.0 due to rounding. -




Table 35

The Total Role Counts by Race for All Subjects
(Percentages) *


Total Role Count


Race


Total


Wh ite+
(N=74)


1.4
(1.4)


0.0 6.8
(1.4) (8.2)


23.0


28.4


18.9


10.8


100.1


9.0) (42.0) (70.4) (89.3)(100.0)(100.0)(100.1)


Black
(N=57)


0.0
(0.0)


0.0 3.5 19.3 13.3 29.8 24.6 1.8 1.8 100.0
(0.0) (3.5) (22.8) (42. 1) (71.9) (96.4) (98.2) (100.0) (100.0)


X = 0.65 af = 8


0 '~i>


Race











significant chi square statistic and the data, on being examined,


failed to show


a demonstrable difference.


The results of the final test (perceived lifespace)


are de-


picted in Table


These distributions also proved not to


be statistically


. significant at the acceptable level of confidence and the


data


within the


table do not


appear


to indicate


a clear pattern.


Table


The Perceived Lifespace


Score


for All


Subjects


(Percentages)*


Total Perceived Lifespace


Score


Race


Total


17-19


White+
(N=74)


14.9
(14.9) (


18.9


13.5


12.2


100.1


.8) (63.5) (77.0) (85.1) (97.3) (98.7)(100.1)(loo.1)(100oo.1)


Black+
(N=57)


28.1


12.3


15.8


100.2


.1) (49.2) (61.5) (77.3) (80.8) (89.6) (94.9) (98.4)(100.2)(100.2)


X = 3.2 df = 8 p > .05
-Parenthetical figures denote cumulati


+Total


percentages


do not


equa


exactyi


e percentage.
00.0 due to rounding.


Since a racial difference


was


not demonstrated


significant variable on the basis of any of the three


as a statistically


measures


engage-


ment, the null hypothesis cannot be rejected.


makes the corollary untenable.


By definition, this also


In addition, none of the three distribu-





77





Because this study constructed formalized hypotheses for test-

ing, it had to be concerned with the statistical significance of the rela-


tionships; based on this

However, a few of the ch


, none of the 13 formulations proved tenable.

i squares calculated were significant and several


configurations of the data appeared to indicate


the existence


of a rela-


tionship.


To summarize


, the results of the hypotheses testing


are:


The hypothesis asserting that being newlywed should


retard the pro


cess


of disengagement


was supported by the


data configurations on all three measure


of engagement,


but none of these


tests


achieved statistical significance.


The hypothesis


asserting that being married should retard


the process of disengagement also appeared to be affirmed


by the configurations of data on all three


engagement, but none of the chi


measures


squares generated from


any of the distributions were of sufficient magnitude


to be significant at an


The hypothes


acceptab


e level of confidence.


asserting that being married to


a spouse


too young to be a subject should retard the disengagement


process of th
configuration


ment (lif


space


older mate


receive


some


supp


of data on two of the measures


measure f


score


ort from the


engage-
fespace


and perceived -i1


measure), but neither test resulted in


a statistically


significant relationship.

The hypothesis asserting that a large number of living
children should tend to retard an aged person's disen-
gagement received partial support because when tested by
the perceived lifespace score, a statistically signifi-


cant relationship


was


the result.


The hypothesis asserting that a large number of living
siblings should tend to retard an aged person's dis-
engagement received partial support, because when tested


by the lifespace measure f
statistic was significant.


score


the resultant ch


square











The hypothesis asserting that the urban elderly should
be less disengaged than the rural elderly received a


degree of support


because


the configuration of data on


the role count test exhibited a tendency toward confirma-
tion, although the variance between the distributions


was


not great enough to produce a statistically signifi-


cant relationship.


The hypothesi


asser


ting that women should


bel


ess


cis-


engaged than men


was categorically disproven


since none


of the three tests generated a significant


statist


none of the data configurations seemed to demonstrate any
observable patterning.


The hypothesis asserting that whites should


engaged than blacks


less dis-


was also categorically disproven


since none of the three tests generated


a significant


statistic and none of the data configuration seemed to


demonstrate any observab


patterning.


With regards to the five corollaries, all have to be considered


disproven since none of their hypotheses were categorically proven


however,


a summary of the consequences of the testing on the corollaries follows:


The corollary which stipulated that the experiment


group subject


with young spouses should be less dis


engaged than comparison group subjects with young


spouses received


some


credence from the role count


measure because that test achieved a statistically


significant difference in the experimental


the variance
statistical


n the comparison group failed
gni ficance.


roup while
to attain


The corollary which stipulated that the experimental


group subject


with the


same


number of


comparison group subjects should be les


acquired partial credence front the


f score since


siblings as
s disengaged


ifespace measure


that test achieved a statistically


significant difference in the experimental group while
the difference in the distributions for the comparison
group did not attain statistical significance.

The corollary which stipulated urban experimental group
^**(? rBe /^* .+ro/ A-/-* I / -,,, I net r* A4 ennn^^B- -jnaAp 4-t-.^-.n *a1rI -,.^ -mn*< r"nrn vk *j-w












The corollary which stipulated that white experimental


group subjects would


less disengaged than white


comparison group subjects also did not attain statisti-


cal significance on


of the three measures of en-


gagement and none of the configurations of data appeared
to demonstrate any observable patterning.


The corollary which


stipulated that the experimental


group subjects with the
parson group subjects


same


number of children


should be less disengaged


as corn


wa s


actually borne out in the direction opposite from that


predicted.


The chi


square tests performed on the dis-


tributions for the perceived lifespace


score


we r e


significant for both groups but the contingency coef-


ficients showed that the relationship


was stronger for


the comparison group.

Of the three measures of engagement used, each produced three


of the nine apparently related configurations of data.


But of the six


tests which


were


statistically significant, the perceived lifespace


measure produced three, the role count measure two, and the lifespace


measure f


score


only one.













CHAPTER FOUR

DISCUSSION


The preceding chapter


sugge is t ed


that although none of the hypotheses


was borne out statistically to an acceptable level of confidence, there


still might be


some


degree


of association between the variables tested.


However


, there may have been


some1


built-in bias that could account for


the differences in dise ngag ment that


we re


noticed between the samples.


Three examples of possible bias are presented.


They are the age differen-


tial between the two groups, differences in the number of other persons


in the household, and differences in life satisfaction.


shown to be statistically significant


statistical significance but ar

This means that the two samples


The first is


The two others did not attain


e shown to be potential sources of b


not compatible on three important


were


variables, and


f a bias


wa S


introduced because of them, it would act


toward confirmation of the hypotheses in the direction predicted.


Hence,


the apparent relationships observed in the data throughout the last

chapter might be the result of these differences between the groups

rather than their variations in disengagement.


Age Differential Between the Groups


is possible that a sample of aged newlyweds might contain more


than its proportionate share of younger old people, that is


persons in











91 Z~r!oi-1t$


proved to


Sn 3CCe


': ISe


of people over 80


(Table 37).


I'


Wi ith more than a third of


I .* S
Ii') I S flU$~Cu
I -


the c~xperi


rental group still in their early o0's and


ncna


of that


.~ C


r cec~mx1


he r 0Oth bi rthdaaywhile only


a fourth of


g (oUp


'h5 .ctr3


4I


early 60's and


a fifth of that


gre I


S.i1


-


years


old; an since


t.ne distributions did differ s-ufwlc h t l y to


bc5 r S.t i 5t i CEiy 1


sign f cent, It must be


assumed


s-Ic <,t


the cc.:risn group d a built-in
.:,e co,.::p rtson group nh dadb i t i


bi as toward d'i sen a~,ement.


T ble 37


Agcs of


and or ,


i.


Si son


C -c~r


for i ~x et n i 1 -- ta
S(Pz _"rcenn .si *


- -- ~ -- ~ .- - - ~ - ~- - -~- .


S; rcul;


Aqc of Res ondents


Total


65--69


70-7 4


73-79


over


Exper ;T..,i .t I
(N = 71)


Conp i
(N ;: So


33.8
(33.3)

25.0
(25.0)


45. 1
(78.9)

28.3
(53.3)


14.1
(93.0)

18.3
(71.6)


7.0
(1GO.O)

8.3
(79.9)


0.0
(100.0)

20.0
(99.9)


100.0
(100.0)

"3.3
(93.9)


-- 1CC5


P-r-.,.het ical. fig, jres denote cumulat ve percentages.
.. .:al percent.ages do not equal exactly 100.0 due to rounding.__


Number of Others in Household

The nurm.'r of other persons in the subject's household is a key


Q~v% i Of


used in two of the three


rfeas~Ures3


of engaceneit (see Appendix A).


fl\i,g


7


~0-;lf











might average out much higher on the number of others in the household


than other noninstitutionalized aged peop


general ly.


This proved to


be the


case


(Tab


Table


Number of Othe


in the Household for the Experi-ental


and Comparison Groups (Percentag


es) ::


Number of Others in House


Group


-i or


Total+


more


Experimental 1
(N = 71)


(8.5)


67.6
(76.1)


(98.7)


(100.1)


100.1
(103.1)


Compa r son
(N = 60)


(23.3)


50.0
(73.3)


16.7
(90.0)


00.1


(91 .7)


(100.1)


4..ePa ren t he2t i ca f igu


+Tota


percentages


res


denote cumulative


do not


equa


1 exactly


percentages.
00.0 due to


round no.


The findings for the comparison group compare favorably with the


national


averages


which show that


percent of the people


over


65 live


alone, 49.5 percent live with


one other in the house, and


27.5


percent


live with two or more other peop


e in the house (Aging, 13:1 and 16:20)


These figures also closely parallel those of Great Britian


where


22.2


per-


cent of the population over 65 live alone and 77.8 percent live with one

or more other persons in the house (Tunstall, 1966:48).

In contrast, the experimental group data showed 76.1 percent living


los












having two or more otlier persons living in the louse with them (26.7

percent for the comparisonrt group).


The fact that mnore than half again


as many of the experiment


group than the co:cprison group


were


living


as couples


was offset by t he


fact


that


;-ivc


time


as many of the comparison


group


subjects shared


their households


with four or more people.


Hence, on balance, it appears


that the assumption that


agedI


new,


ywed s


might genera 1ly have


a greater


number of other people in thie house than aged citizens living in various


other household arrangement


does


seem


to l0old up.


However, the big difference


was the


zero


column, which figured


proportionately larger in the role count tally.


comparison group living alone


The percentage of the


was nearly three times greater than the


percentage of the experimental group living alone, simple


peCatise


spouse


was counted


as another person in the household.


1EI. Ce,


t:;lougrh


the difference between the two groups


as to the number of others in the


house


was not statistically


gnificant,


it is


ear that there


was a


real variation noticeable


e, which must lead us to conclude that a dispro-


portionate number of the comparison group would rank low in the lifespace


measure f


score


and role count measures of engagement.


Results of


a Life Satisfaction Index


at least conceivable that


aged people


who undertake a new


marriage might have more vigor and vitality, and therefore be happier


1~ S1


- -~ -.- 1 1..A ~ 1~ ~ -


. 4. -


"I" ~


Iin 1~ I













zestful senior citizens tend to contract new marriages


and then


becoc,;e


even more happy


as a result


of it.


At any rate, the life


sat isfction


entity must be measured


a against


both groups if for no other


reason


than


to ascertain if it


can be excluded


as possible


cause


of bLas.


To deter-


mine if there


was a variance in satisfaction with life, both


groups


were


given the Life Satisfaction Index A (LSIA), as developed by iNeugarten


et al


(s ee


Appendix A).


The LSIA is divided into five subsections.


The first subject ion,


containing six items


, attempts to


mieasu re


mood tone.


scconI


I


section


, also containing


items, attempts to measure


zest


for living.


The third subscct ion,


Cons


ng of three items, attempts to


measure


congruence between desired


achieved goals.


The fourth subsection,


also consisting of three items, attempts to measure

The lest subsection, of only two items, attempts to


persona

measure


fort i tud,.


the subject's


self-concept


The total of all these subscales is


sunC pOSt.d


to reflect


a respondent's degree of life satisfaction.


Each question


on'the schedule


scored


one


point.


(See scoring key in the Appendix.)


This made


possible range for the mood tone subtota


f rom


zero


six points, the


zest


for life subtotal from


zero


to six, the congruence subtotal from


zero to three, the fortitude subtotal from zero to three, and the self-


concept subtotal from


zero


to two points.


This makes the maximum possibi


range for the total LSIA from


zero


to twenty points.


II r *p --- -IS*a


was











acceptable level of confidence.


Ho'..ver, there


.were


visi ble variations


in the data in


soma~l


of the distributions.


The mood tone results showed


that


percent of the comparison group fell into the two lower scored


columns in contrast to only 14.1 percent of the experin:en tal group while


39.5


parcer;t


of the exper imrental group scored in th- two higher columns


and only 26.6 percent of the comparison group reached those.columns

(Table 39).


Table


Mood Tone Subtotal of the LSIA for the Experimental


and Comparison Group


s (Percen


tact


- .- - -~- -"~~LJ -,,~ ~: ~- --. -,.- -- --- --


Pood Tone Subtota


Group


Total+


Exper iia-n.ta 1
(i = 71)


2.8 11 3 15.5
(2.8) (14.1) (29.6)


11 3 19.7 25.h
,0.9) (60.6) (86.0)


1 0i .
(o100.1)


100.1
(100. 1)


Comperi son
(N = 60)


6.7 18.3 13.3 18.3 16.7 18.3
(6.7) (25.0) (38.3) (56.6) (73.3) (v1.6)


8.3
(99.9)


99.9
(959.9)


*-'Parenthetical figures denotec cumulative percentages.
+Total percentages do not equal exactly 100.0 due to rounding.


The zest for life subtotal also demonstrated an observable dif-


fcrence


s since


more than twice


as many comparison


as experimental subjects


fell in the two lower scored colunnsi while 14.1 percent of the experinmn-r


tal subjects made the highest


score


column in contrast to only 8.3 percent


of the comparison subjects beinG positioned there (Table 40).










Table 40

Zest for Life Subtotal of the LSIA for the Experimental
and Comparison Groups (Percentages)::


Zest for Life Subtota


Group


Total+


Experimental
(N = 71)


2.8
(2.8)


4.2 19.7 23.9 21.1 !4.1
(7.0) (26.7) (50.6) (71.7) (85.8)


' i i.1


99.D


Compa ri son
(N = 60)


6.7 8.3 20.0 23.3 18.3 15.0
(6.7) (15.6) (35.0) (58.3) (76.6) (91.6)


8.3
(S93.)


99.-9
(99.9)


*Parenthetical figures denote cumulative percentages.
--+Total percentages do not equal exactly 100.0 due to rounding.


The variances in the distributions on the congruence were not


1~


but the comparison group did have rore than twice as many respondents in

the lowest scored column than the experimental group and had fewer re-


spondents placed in the highest


scored


column than the experimental group


(Table 4,1).


Table 4


Congruence Subtotal of the LSIA for the Experimental
and Comparison Groups (Percentages)*


Group
0


Congruence Subtota


Total


Experimental+
(N=71)


21.1


(4.2)


(25.3)


36.6
(61.9)


38.0
(99.9)


q99.
(99.9)











The fortitude subtotal displayed a reversal in the trend in that


half again


as many experimental


as comparison subjects fell in the lowest


scored column while more than double the


amount t


of comparison subiocts


than expe r i ren tu subjects fell in


the highest scored column (Table 4\2).


Table 42

Fortitude Subtotal of the LSIA for the Experimental and
Comparison Groups (Percentages)*

- --


Group


Fortitude Subtota
---- --


Total


Expe r i men ta 1
(N = 71)


Comparison
(N = 60)


38.0
(38.0)

21.7
(21.7)


36.6
(74.6)

45.0
(66.7)


21. 1
(95.7)

23.3
(9o0.0)


(399.)

10.0
(100.0)


99.9
(99.9)

100.0
(10l0.0)


*Parenthet ica figures denote


+-Total p onae


do not equal


cumulat


'"'C


exact


percentages.
00.0 due to


rr\* tn djr:!ri


The self-concept resulted in rmore comparison than experimental


being placed in the highest scored column


and also in the lowest


as well,


Shence the weight of one balances the weight of the other (Table 43).


Table (3 .

Self-Concept Subtotal of the LSIA for the Experimental
and Comnparison Groups (Percentages)*


Self-Conceot Subtotal


G rouD


Total


Fnr ~r~r 7tli'


40) 1 1


cnn


LrtL j lil. I I *: IU~ II I 4 %. I -J I I.JC t3 S* StJa'


117


II:











Surprisingly, when all of the subsections were


suWmed


into the total


LSIA, there


was no visibly marked difference between the distributions


Table 44


Tota


LSIA


Score


for the Exoerimental and Comparison


Groups (Percentages)*


Total LSIA


Grouo


3-10


1-13


14-16


17-20


Total


Exper i P-lan ta 1
(a = 71)


23.9
(42.2)


(2.8)


31.0
(73.2)


19.7
(92.9)


7.0
(93.9)


99.9
(99.9)


Comrpar i son
( = 60)


(3.4)


15.0
(23.4)


23.3
(6.7)


33.3
(80.0).


13.3
(93.3)


6.7
(ioo~o


00.0


(100.0)


*'Parenthetical figures denote cumulative percentages.
__ Total percentages do not equal exactly 100.0 due to rounding.


ric


the distributions on the total LSIA produced no statistical


or visually significant relationship, it can be dismissed


as a variable


that cold produce a variance between aged newlyweds and other aged


people in general.


However, it should not be discounted that aged newly-


wed s


seem


to have a slightly higher mood tone, zest for life, and more


congruence between desired and achieved goa


s than other aged persons in


general but seem to have a lesser degree of fortitude.


And,


t is within


the realm of possibility that this lack of fortitude caused them to actively


(Ta~lt Itr:)












For purpose: of analysis and to determine if


a difference in life


satisfaction could' possibly have affected the subhypotheses and working


hypothese-s, the


se en


independent variables of those propositions (i.e.,


marital status, nc tber of children, number of siblings, age of spouse,


residence,


sex


race)


we re,


each in turn, held constant and ran


against each of tLh. LSIA subtotais and total.


tionships resulted


Eleven sign ficant rela-


A numbered topical overview of these results follows:


The nL.niber of living children variable produced


a significant


relationship in the experimental group on the mood tone subtotal.


The number of living siblings variable produced


a significant


relationship in the comparison group


on the total LSIA


score.


and 4)


The age of


spouse


variable produced a significant dif-


ference on the mocKd tone subtotal for both the experimental and comparison
groups.


sc'x


prison group on


variable produced


a significant difference in the conm-


self-concept subtotal.


(6, 7, 8,


10, and ll)


r a rc:


variable produced significant


differences in the experimental group


life subtotal, fortitude


on the mood tone subtotal,


subtotal, and the total LSIA score


zest


in both


groups combined


seems


life satisfaction


C. the mood tone subtotal and the fortitude subtotal.

from the above that the most discriminating measure of


was


mood tone and the demographic variable most signifi-


cant in affecting life satisfaction


was


race.


As regards


race,


it appears that Negroes manifest, to


a noticeable


degree,


a somewhat lesser amount of life satisfaction.


The experimental


group contained


black subjects or 54.9 percent, whereas the comparison


nrnrtn hnrt


h 1 rr-fn rpc nnAnrnn c nr


gR~


nprrP rlt