MEMORY TRAINING IN ADULTHOOD: CHANGING MEMORY AND SELF REGULATION By CARLA M. STRICKLAND HUGHES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 7
2017 Carla Marie Strickland Hughes
To Joan Scully and to individuals ever concerned about their cognitive performance
4 ACKNOWLEDGMENTS I am grateful for the generous support that I was freely given during my dissertation journey. I hope to express my gratitude through action by applying lessons learned on this journey to support others throughout my academic and research career. I owe my gratitude to my dissertation committee Dr. Natalie C. Ebner, Dr. Catherine E. Price, and Dr. James Shepperd, and Dr. Robin West for the h elp and guidance they have given me, for challenging and enriching my ideas, and for shaping me into an independent researcher. I am also grateful to Dr. Susan Bluck for her generous p rofessional advice and support and to Dr. Thomas M. Hess and Dr. Dana Ko tter Grhn for their mentorship of my initiation into psychology and aging research. Words cannot fully express my appreciation for the professional and personal support that I received from my dissertation chair, research advisor, mentor, and friend, Dr. Robin L. West. Her tireless support throughout my graduate journey includes only in part helping me set and achieve realistic goals, providing positively framed honest and constructive feedback, modelling professionalism, how to be a principal investigator and how to have a healthy and fulfilling personal life, and consistently championing my research and academic work. This dissertation would not have been possible without her. For their assistance with all stages of this research from pilot testing and participant recruitment and scheduling to data collection and data entry I thank the undergraduate research assistants of the Aging and Memory Lab, including Edward Ballester, Erika Boone, Blair Cacciamani, Sara Charles, Robyn Cotney, Sharon Faur, Rache l Fisher, Nicholas Garcia, Jessica Green, Brooke Layport, Katie Matthews, Mohima Meera, Arianna Schmid, Alessandra Rosales, Michael Yeber, and Nona Zamiri.
5 Although they cannot be acknowledged by name, I am indebted to the many individuals who volunteered their time to participate in this research. I also owe my gratitude to the individuals and groups who opened their doors to my research program and offered physical space and other support for the assessments and training classes, including Sara Lynn McCra Israel, Oak Hammock at the University of Florida, and Alachua County Senior Recreation Center, and the Millhopper Branch of the Alachua County Library System. For their unwavering support and bound less loving kindness, in the form of open ears and strong shoulders shared advice for the future and reminders of the past with anchors in the present gifts of hope, and demonstrations of healthy priorities, I thank my all my friends and family with spe cial thanks owed to Jay, Michael, and Claudia Hughes Natalie Hadad Emily Mroz, Tatiana Orozco, Liz Redford, Katy Kavanaugh, Shantanu Suman, Obinna Mbah, and Jason Borcherding. I gratefully acknowledge the financial support and recognition of my research with awards from the Gerber Developmental Psychology Research Award, the Charles Vincent and Heidi Cole McLaughlin Dissertation Fellowship, the Robert A. Levitt Student Aging Research Awards, the Jacquelin Goldman Foundation, the American Psychological Ass ociation (Division 20), and the Department of Psychology, College of Liberal Arts and Sciences, Graduate Student Council, and Office of Graduate Minority Programs at the University of Florida.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTI ON TO MEMORY TRAINING ................................ ........................... 13 Need for Memory Training ................................ ................................ ...................... 13 Strategy Training for Episodic Memory ................................ ................................ ... 1 4 Impact of Strategy Training ................................ ................................ ..................... 17 Evidence from Meta Analyses ................................ ................................ .......... 17 Evidence from Experimental Paradigms ................................ ........................... 20 2 GENERALIZATION AND TRANSFER OF TRAINING ................................ ............ 25 Generalization ................................ ................................ ................................ ......... 27 Transfer to Untrained Tasks ................................ ................................ ................... 28 Temporal Transfer ................................ ................................ ................................ .. 29 3 SELF REGULATION AND TRAINING ................................ ................................ .... 33 Self Evaluative Beliefs and Training ................................ ................................ ....... 34 Metamemory ................................ ................................ ................................ ........... 38 4 THE PRESENT STUDY ................................ ................................ .......................... 43 Specific Aims ................................ ................................ ................................ .......... 44 Aim 1. Effectiveness of Abbreviated Training Dosage ................................ ...... 44 Aim 2. Impact of Beliefs Focused Training ................................ ....................... 45 Aim 3. Achieving Near Transfer ................................ ................................ ........ 46 Experimental Overview ................................ ................................ ........................... 47 Hypotheses ................................ ................................ ................................ ............. 49 Aim 1 Hypotheses ................................ ................................ ............................ 49 Aim 2 Hypotheses ................................ ................................ ............................ 50 Aim 3 Hypotheses ................................ ................................ ............................ 51 5 METHODS ................................ ................................ ................................ .............. 53 Participants ................................ ................................ ................................ ............. 53
7 Research Design ................................ ................................ ................................ .... 56 Assessment ................................ ................................ ................................ ............ 57 Procedure ................................ ................................ ................................ ......... 57 Measures and Materials ................................ ................................ ................... 59 Training Procedures ................................ ................................ ......................... 72 6 RESULTS ................................ ................................ ................................ ............... 81 Preliminary Result s: Baseline Data by Condition ................................ .................... 81 Aim 1 Results: Effectiveness of Abbreviated Training Dosage ............................... 82 Training Effects for Name Recall Performance ................................ ................ 83 Training Effects for Memory Self Efficacy ................................ ......................... 83 Training Effects for Memory Control Beliefs ................................ ..................... 84 Training Effects for Strategy Use ................................ ................................ ...... 84 Aim 2 Results: Impact of Beliefs Focused Training ................................ ................ 88 Beliefs Focused Training Effects for Name Recall ................................ ........... 88 Beliefs Focused Training Effects for Memory Self Efficacy .............................. 89 Beliefs Focused Training Effects for Memory Control ................................ ...... 90 Beliefs Focused Training Effects for Strategy Use ................................ ........... 90 Aim 3 Results: Near Transfer Effects ................................ ................................ ...... 91 Preliminary Analyses for Transfer Outcomes ................................ ................... 92 Near Transfer Effect for Object Location Recall ................................ ............... 93 Near Transfer Effect for Occupation Name Delayed Recognition .................... 94 Near Transfer Effect for Occupation Name Immediate Recall .......................... 95 Near Transfer Effect for Occupation Name Delayed Recall ............................. 95 7 DISCUSSION ................................ ................................ ................................ ....... 110 Effectiveness of Brief Memory Training Program ................................ .................. 111 Beliefs Focused Versus Strategy Only Training Approaches ............................... 115 Evidence for Near Transfer ................................ ................................ ................... 118 Limitations ................................ ................................ ................................ ............. 122 Conclusion ................................ ................................ ................................ ............ 125 APPENDIX A STRATEGY CHECKLISTS ................................ ................................ ................... 129 B OJBECT LOCATION VISUAL ASSOCIATION MATERIALS ................................ 132 C NAME OCCUPATION VERBAL ASSOCIATION TASK MATERIALS .................. 135 D HEALTH AND MEMORY ENGAGEMENT SURVEY ................................ ............ 137 Pretest Health and Memory Engagement Survey ................................ ................. 137 Posttest Health and Memory Engagement Survey ................................ ............... 139 E HOMEWORK ASSIGNMENT SHEETS ................................ ................................ 141
8 Homework Assignment Sheet for the Strategy Plus Beliefs Condition ................. 141 Homework Assignment Sheet for the Strategy Only Condition ............................. 142 LIST OF REFERENCES ................................ ................................ ............................. 143 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 154
9 LIST OF TABLES Table page 5 1 Week by week overview of assessment and training procedure ........................ 76 5 2 Overview of measures with references ................................ ............................... 77 5 3 List of training elements to enhance self regulation included in each condition 78 5 4 Examples of self regulatory elements in training materials ................................ 79 6 1 Baseline statistics by training condition ................................ .............................. 97 6 2 Means and standard deviations for primary outcomes for trainees and control group at pretest and posttest ................................ ................................ .............. 98 6 3 Effective strategy use for trainees and control group at pretest and posttest ..... 99 6 4 Means and standard deviations for primary outcomes for SO and SB trainees at pretest and posttest ................................ ................................ ...................... 100 6 5 Effective strategy use for SO and SB trainees at pretest and posttest ............. 101 6 6 Means and standard deviations for transfer outcomes by training condition at pretest and posttest ................................ ................................ .......................... 102 6 7 Correlations for pretest and posttest name recall and transfer outcomes ......... 103 B 1 List of objects used in the object location visual association task ..................... 132 C 1 List of occupation name pairs by set ................................ ................................ 135
10 LIST OF FIGURES Figure page 5 1 Example page of to be remembered name face pairs used in the name recall task ................................ ................................ ................................ ..................... 80 6 1 Name recall performance by time for trainees and control participants ............ 104 6 2 Memory self efficacy by time for trainees and control participants ................... 105 6 3 Name recall strategy use by time for trainees and control participants ............. 106 6 4 Memory performance (% correct) on transfer outcomes at pretest for all participants ................................ ................................ ................................ ....... 107 6 5 Occupation name immediate recall performance by condition at pretest and posttest ................................ ................................ ................................ ............. 108 6 6 Occupation name delayed recall performance by condition at pretest and posttest ................................ ................................ ................................ ............. 109 B 1 Example objects used in the object location visual association task ................ 133 B 2 Example matrix array used in object location visual association task ............... 134 C 1 Occupation name verbal association task instructions ................................ ..... 136
11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MEMORY TRAINING IN ADULTHOOD: CHANGING MEMORY AND SELF REGULATION By Carla M. Strickland Hughes August 2017 Chair: Robin Lea West Major: Psychology Everyday memory ability such as learning and remembering names is highly valued by people of all ages, yet is known to decline normatively with increasing age, when related non ability self regulatory factors, such as self evaluative beliefs and effect ive strategy selection, are also at increased risk. Collective results from cognitive interventions demonstrate that memory can be improved at any age, but controversy surrounds the true value of these programs in terms of the scope and duration of trainin g related gains. A traditional and common approach to cognitive interventions for adults is memory strategy training, and limited work of this type has examined whether self regulatory factors might benefit from these programs or moderate other training re lated gains. Further, while interventions focused on intensive practice or core capacity training have demonstrated near transfer after training, performance improves on untrained tasks related to the target task evidence of near transfer from strategy training programs is very rare. The present research addressed both self regulation and transfer issues. Could a short term training program, focused on strategy training, lead to changes in beliefs and strategy usage, as well as score gains? Would a shor t program emphasizing self
12 regulatory change be more effective than a program focused solely on strategy training? Would near transfer be possible if both performance and self regulation changed as a function of training? Overall, this research demonstrat ed that a brief name recall strategy training program is effective for middle aged and older adults. The training program constituted two hours of in person instructor led training and approximately two to three hours of continued self study at home. After training, as compared to pretest data and as compared to inactive waitlist control participants, trainees demonstrated improved name recall memory, higher levels of memory self efficacy, and more effective use of memory strategies. Contrary to expectatio ns, benefits from training were similar for a beliefs focused strategy training approach, as compared to a content and duration matched training approach without the focus on beliefs. However, benefits from both training approaches extended beyond the tar geted name recall task that was trained to performance on similar, but untrained, associative memory tasks.
13 CHAPTER 1 INTRODUCTION TO MEMORY TRAINING Need for Memory Training Older Americans constitute a growing segment of the population, and adults age d 65 years or older today can expect to live another 20 years, on average (Federal Interagency Forum on Aging Related Statistics, 2012). While the quantity of these additional years is obviously desirable, the quality of living during this time may lack pr oductivity and pleasure due to memory difficulties, both subjective and objective. Older adults rate their memory as highly important to their lives (Dark Freudeman, West, & Viverito, 2006), and independent living requires a certain degree of memory abilit y (Berry, Hastings, West, Lee, & Cavanaugh, 2 010; Hering, Rendell, Rose, Schnitzspahn, & Kliegel, 2014; Stine Morrow & Basak, 2011). At the same time, extensive research shows that older adults show normative declines across age in several aspects of memor y (Berry et al. 2010; Mather, 2010; Nyberg, Lvdn, Riklund, Lindenberger, & Bckman, 2012). In older adults, complaints about such memory changes are related to lowered ability to perform activities of daily living and reduced quality of life (Montegjo, Montenergo, Fernndez, & Maest, 2012). Declines in memory (e.g., forgetting names) may determinately impact the quality of social interactions, as well as work and/or volunteer experiences Older adults, compared to younger adults, are also less likely to successfully employ memory strategies (McDaniel & Bugg, 2012). At the same time, negative stereotypes about aging and memory are pervasive, creating fear of more significant memory loss (Dark Freudeman et al., 2006) These age changes in memory are proble matic and worrisome, although not insurmountable.
14 Increased research evidence supports the existence of cognitive plasticity in late life, challenging stereotypes concerning the assumed inevitability of memory decline with age. This plasticity enables per formance improvement, even in the very old (Hertzog, Kramer, Wilson, & Lindenberger, 2009; Stine Morrow & Basak, 2011). Indeed, a long accepted body of work has established that older adults may benefit from memory training (Berry et al., 2010; Gross et al ., 2012), although, the practical impact of these programs is less well established Traditional memory training programs demonstrate little transfer of benefit to untrained tasks, longer term benefits are uncertain, and some memory strategies may be diffi cult to apply in everyday life (Rebok, Carlson, & Langbaum, 2007; West & Strickland Hughes, 2015). One possible explanation is that most programs do not address negative beliefs about memory and aging, which can discourage older adults from investing long term effort in the memory strategies learning during training (West, Welch, & Yassuda, 2000). The present project addresses this issue: by examining an intervention designed to maximize improvement in memory beliefs as well as strategies and memory perfo rmance Before describing the methods for the present project, this paper presents an overview of the extant literature on memory training and aging, with a focus on strategy training, the most common approach to aging and memory intervention. Strategy Training for Episodic Memory N umerous cognitive interventions over the past half century have targeted memory improvement in the aged. On the most basic level, these interventions are effective; immediately following training, older trainees perform better on the trained memory tasks than they did before training, whereas control participants do not demonstrate comparable improvements (G ross et al., 2012; Hertzog et al., 2009; West,
15 2012). At the same time, the promise of improved outcomes via memory training needs to be evaluated in relation to the degree to which everyday memory skills can be improved via training. At present, the overa ll effectiveness of training varies dramatically across programs that differ along multiple dimensions, such as their design, format, duration, or content. For example, memory training programs have been designed to target core processes, such as processin g speed or working memory (via extended practice), to enhance overall cognitive engagement with lifestyle changes, or to provide instruction on memory strategies. The formats of training programs include both in person classes or training and self study pr ograms to complete at home. Program content might be delivered via paper and pencil or technology such as video recordings, computers, or video games. Participation in these programs may last a few hours or a few months. Some programs train a single memory strategy or practice a single activity; multifactorial programs include multiple types of activities and may supplement strategy content with information about normative aging and memory changes or practice in non strategy techniques, such as focused atte ntion (Hering et al., 2014; Rebok et al. 2007; Stine Morrow, Parisi, Morrow, Greene, & Park, 2007; Zehnder, Martin, Altgassen, & Clare, 2009) Training benefits are examined by comparing performance levels before and after a specific memory intervention. However, observed pre post improvements may be credited to factors other than the memory intervention. For example, performance commonly improves from repeated testing or additional exposure to specific tests and measures, gains which are known as practice effects (Ball et al., 2002; Hertzog et al., 2009). Further, performance gains could result from aspects of the intervention other
16 than the active ingredient (i.e., practice or strategy training), such as social interaction or changes in motivation and out come expectations (Boot, Simons, Stothard, & Stutts, 2013; Stine Morrow & Basak, 2011). To address these alternative explanations for pre post gains and to test for specific training related performance improvements, pre post performance of trainees can be compared to that of control groups. Control groups complete the same assessments on the same schedule as the trainees but do not complete the target training or intervention activities. In such designs, the effect of training is represented by pre post im provements specific to trainees, above and beyond any gains demonstrated by control groups. Random assignment of participants to training or control conditions, such as in randomized control trials, establishes comparable groups representing the population of interest (Boot et al., 2013). Without an appropriate control group, the internal validity of an intervention study is suspect and interpretation of pre post gains is limited. Control groups may be inactive or active. An inactive control group typicall y completes the same interviews as trainees, but has no training meetings or other activities. Waitlist control groups are inactive, but are scheduled to complete the training program following all assessments. Comparison to inactive/waitlist groups contro ls for effects of practice from repeated testing. Active control groups will hold meetings and/or participate in activities that are distinct from the target training. Their sessions are matched to training along dimensions related to time, number of meeti ngs, and/or opportunities for social interaction. In this way, comparison to active control groups can further control for level or type of activity, social experience, motivation, and other aspects of participation such as the laboratory environment or tr
17 performance gains. For example, compared to memory strategy classes, botany lectures may be matched in instruction style, difficulty of the course material, and duration and frequency of class meetings. Active control groups may be placebo or non placebo. Placebo groups are active but participate in activities which should not benefit or otherwise influence the outcome of interest. Non placebo control groups participate in a different type of training program (e.g., exercise versus strategy classes), different versions of a training program (e.g., self study versus instructor led), or different training approaches or formats (Bailey, Dunlosky, & Hertzog, 2010; Dunlosky, Kubat Silman, & Hertzog, 2003; Hastings & West, 2009; Stine Morrow et al ., 2014). Impact of Strategy Training Cognitive interventions target memory improvement for middle aged and older adults because episodic memory (e.g., learning new names, recalling a story) is highly valued and fundamental to daily life, yet declin es wit h increased age (Dark Freudeman et al., 2006). Accordingly, the success of such an intervention can be judged by whether it reliably enhances memory performance. Fortunately, over half a century of research demonstrates these programs are successful using this metric: episodic memory performance of middle aged and older adults is typically better following training than before training, even when accounting for practice effects with appropriate control groups (Berry et al., 2010; Gross et al., 2012; Stine M orrow & Basak, 2010). Evidence from Meta Analyses Meta analyses establish that strategy training programs for older adults are successful in enhancing memory performance. The meta analyses reviewed here focused on memory and mnemonic training. However, not able evidence from other
18 systematic reviews on cognitive interventions suggests mature adults may benefit from computerized and video game interventions (Kueider, Parisi, Gross, & Rebok, 2012; Toril, Reales, & Ballesteros, 2014) and training core capacitie s, such as attention or working memory (Morrison & Chein, 2011). For memory, meta analyses have examined training effects relative to inactive control groups or active placebo groups, or included studies without control groups, as noted below. An early me ta analysis of mnemonic training for older adults confirmed greater gains for trainees compared to active and non active control groups (Verhaeghen, Marcoen, & Goossens, 1992). As an initial examination of the overall potential of mnemonic training, this r eport focused on pretest to posttest comparisons but included training programs which did not employ control groups. Overall pretest to posttest change was greater for trainees than for the available control participants. In addition, gains were greater fo r programs using group training, rather than individual training. Interventions incorporating pretraining, such as relaxation training or other non mnemonic instruction, showed higher gains than those without pretraining. However, gains were unaffected by the type of strategy trained or the content of the pretraining (Verhaeghen et al., 1992). These principal findings regarding episodic memory benefit from intervention have been replicated. With colleagues, Gross completed a meta analysis involving 35 studi es that examined memory training conditions for older adults (without pharmaceuticals or other intervention, such as exercise), in comparison to control conditions (Gross et al., 2012). The overall effect of training across these studies was significant: o bjective memory performance improved across assessments more for individuals who received training
19 than for those who did not complete training. As in the previous meta analysis, the impact of training was unaffected by the particular strategies that were trained. Further, the report documented that an intervention offering only one strategy typically shows weaker effects than an intervention which provides training on many strategies. A review of cognitive intervention programs for older adults conducted by Zehnder and colleagues (2009) reported less promising findings than those reported by other reviews. They considered programs for adults with mild cognitive impairment, as well as for healthy older adults. Similar to the previous meta analysis (Gross et al., 2012), pre post improvements in episodic memory performance were greater for trainees than for inactive control participants across the reviewed studies. In particular, positive impact of mnemonics training was found for paired associates learning an d verbal recall. However, pre post improvements in memory were not superior for trainees when compared to active control groups. One possible interpretation is that non specific training effects contributed to memory gains, as pretest to posttest compariso n of memory performance was similar for memory trainees and active control groups. However, Zehnder and colleagues broadly defined active control groups, including alternative interventions (e.g., relaxation) and combined programs in which memory was train ed along with other intervention activities (e.g., exercise). Thus, their supposition that memory gains from strategies training may not exceed that derived from active control groups should be interpreted with considerable caution. Collective evidence f rom meta analyses confirm that memory training for older message s are many. Training is effective for people who are over
20 60 years of age. Participation in cognitive interventions is better than doing no intervention activity. More training, with varied activities, is better than less training and with fewer types of exercises, and group training programs are more effective t han individual programs. It is also important to note that the benefits to episodic memory from training programs may be derived from aspects of the interventions that are not specific to the targeted training activities, considering that the effectiveness of these programs does not seem to vary in relation to the particular strategies that are trained. Evidence from specific experimental training programs can be used to document the global effects reported in these meta analyses. Evidence from Experimental Paradigms Reports from experimental studies provide additional understanding of the memory enhancement that comes from cognitive training. As mentioned, intervention designs vary greatly In this section, several experimental studies are reviewed as examp les t o demonstrate the fact that existing interventions vary greatly in length and approach to training. In particular, these studies highlight the potential for memory gain from varied programs. ACTIVE trial The Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) trial was a randomized clinical trial for older adults. The project was large scale, including over 2,800 participants Training was completed in six U.S. cities using a single blind approach (Ball et al., 2002; Rebok et al., 20 14; Willis et al., 2006). Participants completed a large battery of cognitive tests and surveys several times, and assessments continued over ten years. The ACTIVE trial compared three different training conditions. In this way, each training group served as an active, non placebo
21 control group for the other training groups. Memory training and associated memory performance outcomes are of interest here. Memory trainees ( n = 711) attended ten classes, for about an hour each week. During each small group ses sion, participants learned or practiced strategies, such as forming mental associations or visual images, received feedback, and completed practice tests varying in ecological validity. Strategies were introduced in the first five sessions and were further practiced in later sessions. Thus, the memory training program of the ACTIVE trial extensively trained multiple strategies. The target ability for memory training was verbal episodic memory (e.g., list or story recall; Ball et al., 2002). Overall, the ACT IVE intervention resulted in improvements in episodic verbal memory for older adults and seemed to attenuate expected decline over time due to normal aging. However, the observed benefits for memory trainees were considerably less robust than trainees in t he reasoning or speed of processing groups (Ball et al., 2002; Ball, Edwards, & Ross, 2007; Rebok et al., 2014 ; Willis et al., 2006 ). More than one fourth of the intervention participants showed significant change in memory scores after training. Performan ce on the verbal memory tasks was better for memory trainees than the other groups one year and two years following training ( Ball et al., 2002). In contrast, memory training related benefits for verbatim story recall were transitory, as they were demonstr ated immediately after training but were diminished at the one year follow up (Sisco, Marsiske, Gross, & Rebok, 2013). Maintenance of training related gains is discussed in detail later. Using a sample of older individuals with subjective memory complaints Cohen Mansfield and colleagues (2014) replicated the ACTIVE memory training progr am, with
22 minor revisions, in another randomized control study. In this study, benefits from the ACTIVE memory program were directly compared to benefits from two comparable cognitive intervention programs one that focused just on participation and one that focused on general health and wellness. The health promotion program included structured classes and homework on topics related to health behaviors and aging. The partici pation centered course was formatted like a book club on memory strategies and promoted goal setting and supportive group interactions. All participants completed a baseline assessment before training. The timing of assessments and training was staggered s uch that some intervention groups served as waitlist comparison groups for others who received training earlier. Cognitive functioning was assessed objectively with a battery of assessments that included measures of episodic memory (verbal and non verbal). Importantly, cognitive functioning was better following the training program for all groups. While all groups improved objectively on the global cognition score, the ACTIVE training group, but not the other groups, reported reduced subjective memory compl aints representing Subjective memory gains may be an important antecedent for objective performance improvement (Crumley, Stetler, & Horhota, 2014), as discussed in more detail later. The subjecti ve memory gains exclusive to the ACTIVE training group may be explained by the relatively greater face validity and social validity of this condition compared to the others. Overall, the data from the ACTIVE study, and its replication, revealed gains in be liefs about ability as well as changes in episodic memory performance
23 EMC project Another exemplar cognitive intervention is the Everyday Memory Clinic (EMC) project (West, Bagwell, & Dark Freudeman, 2008) which focused on training strategies for severa l types of episodic memory tasks. Participation in this training program lasted five weeks, and the program was multifactorial, in that trainees completed several different types of exercises and activities, above and beyond strategy instruction These inc luded attention training a s well as readings on normative age related cognitive changes Adults over 50 years old trained on multiple mnemonics through weekly group meetings, homework and practice (in class and at home). EMC is most remarkable for its fo cus on improvement in self efficacy vehicle for training related memory gains (West et al., 2008; West & Hastings, 2011). All training a ctivities were designed y. Episodic memory performance (name, list, and story recall) and memory beliefs were evaluated on three occasions, before and after training, and one month later. Compared to the performance of a waitlist control group, performance of the trainees improve d after training and at follow up testing one month later, for both name recall and story recall. Robust training effects were not apparent for list recall. List recall is relatively less challenging than the other two recall tasks, and supportive external aids are effective and easy to implement for this particular memory demand (Mather, 2010; McDaniel & Bugg, 2012). EMC trainees were perhaps more motivated to focus on improving the more difficult tasks, and control participants may have performed well on list recall tasks without training. In addition to improved name and story recall performance, strategy use was more effective for trainees compared to control participants, as indicated by self report
24 strategy lists following each memory task (West et al. 2008). Training gains were with intervention assignments and quantity and qual ity of participation. Further, a follow up report evaluated benefits from a self help training condition wherein trainees completed a workbook version of the EMC intervention at home (Hastings & West, 2009). Recall memory improvement was greater for the se lf trained group than for the non active control group. In sum, cognitive interventions for older adults strive to enhance episodic memory, as a potential key to prolonged independent living and enhanced quality of life. Collectively, as reported in meta a nalyses, training interventions can successfully enhance memory performance. In addition, the experimental evidence from specific intervention programs documents the potential benefits of strategy training. Both the ACTIVE trial and the EMC program trained groups of older adults on multiple mnemonics over fairly long periods. The ACTIVE trial was a traditional, yet extensive, intervention focused on training strategies. The EMC program was multifactorial and also included a novel focus on self efficacy (dis cussed in more detail later). Importantly, the success of these programs document the potential for episodic memory gain via mnemonics training.
25 CHAPTER 2 GENERALIZATION AND TRANSFER OF TRAINING T he primary concern of memory training is whether it may effectively enhance memory performance. However, the practical impact of memory training for older adults is potentially much broader. In particular, generalization and transfer of benefit are important mechanisms for creating broader impact. Generalization occurs when training provides benefits for broader cognitive, social emotional, or physical abilities, promotes engaged, healthy life styles, or improves general well being (Hertzog et al., 2009; Stine Morrow & Basak, 2011). More limited t ransfer effects are specific training related gains in non trained cognitive tasks. Barnett and Ceci (2002) recommend classifying transfer effects acc ording to the content of the learned and transfer skills (e.g., level of specificity or similarity of trained tasks to transfer tasks ) and the context of transfer (e.g., time since training). With respect to specificity, n ear transfer effects are improveme nts in performance on tasks closely related to the trained ability, such as gains in visuospatial working memory when verbal working memory is trained. Far transfer effects are improvements in performance on tasks which are less closely related to the trai ned ability, such as enhanced reasoning ability after episodic memory training (Zelinski, 2009). Transfer can also vary along a temporal dimension, with near transfer reflecting benefits within a week or month of training (West et al., 2008) and far tempor al transfer reflecting gains that can be maintained over months or even years (Rebok at el., 2013). The practical impact of specific training programs, in terms of both types of transfer, is controversial. Some scholars argue that broad generalization of b enefits from specific training programs is assumed, despite limited evidence for near or far
26 (McDaniel & Bugg, 2012; Salthouse, 2006). However, (cf., Fish er, 2012). Failure to find transfer is often reported in training reviews even though the benefits of cognitive intervention may extend beyond specifically trained memory tasks (Hertzog & Dunlosky, 2012; Kliegel & Brki, 201 2 ; Zelinski, 2012). More importa ntly, much of the variance in the research evidence for transfer may be related to inconsistencies in the definition and assessment of transfer (Fisher, 2012; Zelinski, 2009). Regardless of the categorization of generalization and transfer effects, assessm ent and explanation of these benefits are integral to maximizing the practical impact of cognitive interventions. Generalization and transfer are important because they increase the value (e.g., benefit per investment in training) of interventions and impr ove the daily lives of trainees (Barnett & Ceci, 2002; Zelinski, 2009). In the specific context of cognition and aging, benefits to non trained abilities may stave off age related attenuation in cognition (Zelinski, 2009). Therefore, generalization and tra nsfer are an important goal of memory interventions for older individuals. Generalization and transfer are commonly assessed via a comparison of outcomes of interest (e.g., survey responses, task performance) between trainees and control participants. The first step is to identify scores for cognitive performance on the non trained tasks or reported levels for general aspects of function (e.g., depression, neurogenesis) in both groups. Comparison to control groups is essential for understanding the magnitud e of change. For example, maintenance, rather than improvement, for trainees may be consequential if control participants demonstrate deterioration in a targeted ability (Hertzog et al., 2009). Utilizing active comparison
27 groups who participate in alternat ive training programs may further allow for extensive evaluation of transfer effects, if the targeted training outcomes differ between the intervention groups. For example, the ACTIVE trial compared performance on a battery of cognitive tests across groups of older adults receiving focused training in memory, reasoning, or speed of processing (Ball et al., 2002; Rebok et al., 2014). Thus, memory performance represented a direct training effect for the memory group but a transfer effect for the reasoning and speed of processing groups. Generalization Looking at the evidence for generalization, diverse outcomes may profit from cognitive training and these outcomes may protect against normal age related declines. Generalization has been demonstrated to some no n cognitive outcomes such as social support (Fried et al., 2004 ), increased neural activation (Carlson et al., 2009), daily function (Rebok et al., 2014; Willis et al., 2006), and other indicators of well being and health (Cohen Mansfield et al., 2014; Fri ed et al., 2004). For example, cognitive interventions have led to changes in feelings of loneliness or depression, (Cohen Mansfield et al., 2014), and the latter could be a risk factor for dementia (Ownby, Crocco, Acevedo, Vineeth, & Loewenstein, 2006). B rain aging (e.g., increased white matter abnormalities, dysfunction such as over and under recruitment, decreased volume and grey matter) is also hypothesized to explain normal cognitive decline with aging (Hayes & Cabeza, 2008). With increased evidence o f neuroplasticity, even in late life, engaged lifestyles and participation in cognitive interventions may promote neurogenesis (Park & Bischof, 2013). Ultimately, the benefits from cognitive interventions may broadly generalize to overall health and well b eing. Participation in the ACTIVE trial predicted improved functionality and ability to perform activities of daily
28 living after five years, although these effects were not evident immediately following training (Rebok et al., 2014; Willis et al., 2006). N otwithstanding this evidence for generalization, there are relatively few studies focused on broad generalization of training effects and considerably more work needs to be done. Transfer to Untrained Tasks Th e evidence suggestive of broad generalization o f training is complemented by evidence concerning specific transfer to non trained cognitive tasks. Successful transfer is evident when core capabilities are trained; these training paradigms involve intensive practice (e.g., 20 or more hour long training sessions over several weeks) of specific information processing skills, such as working memory or visual attention (Zelinski, 2012). As compared to control groups, trainees with extensive working memory practice often exhibit gains in higher order cognitiv e processes, such as reasoning and composite scores of general fluid intelligence (Morrison & Chein, 2011). Interestingly, these gains in fluid intelligence from core capacity training are dose dependent, with increased benefit from longer or more intensiv e training programs (Jaeggi, Buschkuehl, Jonides, & Perrig, 2008). Cognitive training via computer or video games represents a comparable training paradigm to extensive practice of core capacities in that important cognitive processes (e.g., focused attent ion, processing speed, working memory) are well exercised by video game play (Hertzog et al., 2009). Video game training is effective for older adults; across 20 studies analyzed in a meta analysis, pre post improvement in several cognitive performance out comes (e.g., memory, reaction time, attention) was greater for video game trainees than control participants (Toril et al., 2014).
29 Thus, s everal different training paradigms appear to promote transfer effects, particularly for general cognition or fluid i ntelligence, but transfer from strategy training on one type of memory task to another type of memory task is rare (Charness, 2007; Fisher, 2012). In the future, it would be important to focus on the specific conditions that promote transfer, that is, the how and when of transfer, particularly those conditions that make it possible for trainees to improve their everyday memory function (West, 2012). Temporal Transfer Temporal transfer represents gains from interventions that persist over time. Cognitive interventions proposes that a single training dose should sufficiently protect against eflects the impact of cognitive interventions, Hertzog and colleagues (2009) propose: One shot inoculations, however intensive, may not be as effective as creating contexts in which newly trained procedures are reinforced by contact with trainers and other s, communication about challenges to implementing the procedures in everyday life, and instruction in how to overcome those challenges by adapting trained procedures to contextual variations. (p. 45) That is, lasting gains from cognitive interventions are likely promoted by continued cognitive exercise and related behavior changes, as in a physical intervention. Sparse empirical evidence tests these alternative models of temporal transfer in cognitive interventions. L ittle is known about temporal transfer, or how long the benefits of training are maintained or sustained. Few intervention studies test and report long
30 term training outcomes; most programs report only immediate training gains or those within one month or a year of training (Gross et al., 2012) Thus, documenting temporal transfer would be critical for establishing the practical value of training. Evidence from some studies suggests a regression to baseline performance following initial training related performance boosts. For example, comparis on of pretest memory with performance three years following strategy training revealed declines in everyday memory performance (specifically, face name associations and grocery store shopping lists) for trainees, compared to no significant change for the i nactive control groups (Scogin & Bienas, 1988). That is, trainees did not maintain their initial training gain, as their performance returned to pretraining level. In another study, maintenance of memory gains was assessed two years following training of memory strategies (11 sessions, 90 minutes each), using a battery of everyday memory tasks and self report measures of related beliefs (Bottiroli, Cavallini, & Vecchi, 2008). As before, trainees exhibited initial gains from training, but these gains were l argely diminished two years later, when performance was comparable to initial levels. Exceptionally, gains in face name associative memory were maintained over the two years. This specific long term maintenance is noteworthy because face name associative m emory is highly challenging, especially with advancing age (Strickland Hughes, West, Smith, & Ebner, 2016). As with the maintained face name associative memory performance, a few other interventions demonstrated lasting memory gains beyond a year following training. Stigsdotter Neely and Bckman (1993) report long term follow up data for two of their intervention studies. In particular, they evaluated maintenance of training effects for 3.5
31 years. Both studies compared an 8 week long, multifactorial trainin g program focused on encoding strategies, attention, and relaxation to inactive control groups and alternative training programs. Participants in the multifactorial training conditions exhibited relatively better recall memory than the other groups. In one study, multifactorial trainees demonstrated improved performance at 3.5 years compared to 6 months after training. The other two groups maintained but did not improve their performance level over this time. In the other study, participants in each of the three groups demonstrated no change in performance from 6 months to 3.5 years following training. However, performance was greater for the multifactorial trainees than the other two groups at the 6 month posttest. Thus, some interventions demonstrated main tained memory gains but others did not. Importantly, each of these studies examining long term impact had small sample sizes and may have been under powered. T he first long term assessments of training benefits with a large sample were reported by the ACTI VE trial, which tested over 2,800 participants up to ten years after training (see 1 and 2 year data in Evidence from Experimental Paradigms ). Participant retention rates following training were 67% after a 5 year d elay (Willis et al., 2006) and 44% after a 10 year delay (Rebok et al., 2014). Analyses were conducted using intent to treat procedures. Memory performance was improved five years following training for trainees from all conditions (memory, speed, and reas oning training), as compared to performance before training (effect size for difference in untrained control group was 0.23 SD; Willis et al., 2006). Latent growth curve analyses examining the slope of change in a memory composite score further indicated memory performance may improve through year five (Rebok et al., 2013). Interestingly, this gain was unaffected by
32 up training intended to promote maintenance of gains) that were completed by a subset of trainees one year after initial training. While maintenance of gains through year five was suggested, neither the memory training nor the booster sessions were related to significant performance gains one decade after training. (Rebok et al., 2014). The initial lar ge scale, long term evidence for training benefits from ACTIVE are promising, but additional longitudinal research is warranted to understand the potential for temporal transfer from cognitive interventions. In sum, the collective evidence from cognitive interventions show great promise in extending benefits to the everyday life of older adults. Generalization has been evidenced in that non cognitive tasks may benefit from training regimens, and, importantly, these outcomes are often related to cognitive h ealth. Transfer to cognitive tasks, such as attention and fluid intelligence, is also evidenced by training programs (Hertzog et al., 2009; Toril et al., 2014). Long term transfer over 5 years has also been established. Thus, benefits from cognitive interv entions show specific transfer and broadly extend beyond specifically trained memory tasks to practical and relevant skills, abilities, and outcomes. In particular, one benefit of cognitive interventions may be enhancement of self regulation. This evidence will be considered in the next chapter.
33 CHAPTER 3 SELF REGULATION AND TRAINING While improved performance is the ultimate goal of memory training programs, the role of self regulatory factors in intervention is of increasing importance to clinicians and scholars of cognitive aging. The social cognitive theory of self regulation proposes that personal, behavioral, and environmental factors, as well as their interactions, collectively determine performance above and beyond individual ability factors (Bandu ra, 1997 ; Nelson & Narens, 1990 ). Self regulation occurs when individuals responsively and effectively adapt after evaluating the personal and environmental factors relevant to a particular challenge. Thus, successful self regulation involves maximizing o related actions. Self regulation, in the context of memory, can be seen as involving strategy selection (e.g., pick cognitive tasks. In a self regulatory context, self evaluative beliefs (judgment or appraisal of self an d individual capabilities ) and m etamemory (knowledge about memory and memory strategies ) are important elements of self regulation. Consider two examples which respectively demonstrate the self regulatory influence of self evaluation and metamemory In the first increases after he receives positive feedback from practice levels of a memory game (Strickland Hughes et al., 2016 ). This more positive self evaluation promotes sustained effort during subseq uent memory challenges, and he is more likely to attempt harder
34 game levels. In the second example, w hen faced with the challenge to remember names of new acquaintances at a party, an older woman considers the mnemonics she knows and recruits the most suit able strategy (Hertzog & Dunlosky, 2012). As she continually judges how well she has learned each new name, she invests less effort in practicing the names she has already learned and more effort in learning the ones she examples highlight the potential for self regulatory processes to maximize memory outcomes. Self Evaluative Beliefs and Training Memory related self evaluative beliefs are central to well being, broadly, and to cognitive health, specifically. More positive self evaluative beliefs are related to greater quality of life (Montegjo et al., 2012) and lower levels of depression (Floyd & Scogin, 1997). This self regulatory factor can even predict late life mortality (Wiest, Schz, & Wurm, 2013). More relev ant to the present research, more positive self evaluative beliefs are related to better concurrent and future memory performance (Be a udoin & Desrichard, 2011; Crumley et al., 2014; Valentijn et al., 2006). These documented relationships support the notion that beliefs affect the way in which individuals approach information processing tasks. Age differences in self evaluative beliefs are often reported in the literature (West & Strickland Hughes, 2015) and the link between self evaluation and performance s trengthens with age (Agrigoroaei, Neupert, & Lachman, 2013; Blanchard Fields, Horhota, & Mienaltowski, 2008). Given these relationships, it is likely that self evaluative beliefs play a central role in self regulation during task performance. Two importan t questions regarding self regulation are pertinent to the study of cognitive interventions for the aged. Do cognitive interventions enhance self regulatory
35 factors? To what extent are training related memory gains strengthened or sustained by self regulat ory factors? Research evidence suggests memory related self evaluative beliefs benefit from cognitive interventions. Floyd and Scogin (1997) completed a meta analysis of 27 memory training studies. Their analysis focused on subjective measures, specifical ly metamemory assessments and other indicators of subjective memory, such as performance predictions. With an average participant age of about 70 years old (total N = 1,150), their analysis showed a significant effect ( d = .19). This effect suggested that memory training improves subjective memory measures absolutely At the same time, subjective memory improvements were smaller than improvements observed in objective memory scores, and interventions emphasizing beliefs showed stronger effects on subjective indicators. Later experimental research suggests that memory training may enhance self evaluative beliefs and that these beliefs in turn predict or explain training related gains on cognitive outcomes. Both the ACTIVE study and EMC project, described prev iously, measured change in beliefs. One of these self regulatory beliefs was personal locus of control for memory a sense that one can improve memory through personal effort Memory control beliefs are related to enhanced memory performance and intellectu al function across the lifespan (Agrigoroaei & Lachman, 2010). Although memory control beliefs generally decline with increasing age (Lachman, Neupert, & Agrigoroaei, 2011) they may be modifiable through instruction, even in late life (Hastings & West, 20 11; Lachman, Weaver, Bandura, Elliott, & Lewkowicz, 1992).
36 Level and change in memory control beliefs were measured in the ACTIVE study. When examined five years after training, the likelihood of improved memory control beliefs was considerably greater fo r reasoning and speed of processing trainees than the inactive control group (Wolinsky et al., 2009) despite no particular emphasis on these beliefs in the training program. Unfortunately, this control beliefs change did not extend to the ACTIVE memory tr ainees, although a replication study evidenced trainee improvements in subjective memory (Cohen Mansfield et al., 2014). Personal locus of control for memory is a self regulatory factor that we hope to positively influence in the current study. Training re lated gains in memory self efficacy beliefs are also central to this proposal. Self a specific domain. Bandura (1997) posits that motivated behavior, task persistence and task suc cess are regulated by self efficacy. In turn, self efficacy derives from enactive mastery, vicarious experience, verbal persuasion, and physiological and affective responses. The design of the EMC project included several elements to maximize these four in fluences on self efficacy. For example, self efficacy elements in the EMC project included opportunities to experience in class success remembering names of other class members (enactive mastery), small group discussions of the homework questions (vicariou s experience), provision of positive feedback from the trainer within the classes and on first posttest scores (verbal persuasion), and an emphasis placed on learning, not on scoring 100% (physiologic and affective states; West et al., 2008). One month fol lowing the intervention, EMC trainees reported increased memory control beliefs and greater levels of memory self efficacy, whereas the control participants reported
37 declines in these beliefs. These EMC project results provide stronger evidence of the poss ible benefit of training to beliefs than that reported by the ACTIVE trial. Trainees who completed an at home self study version of the EMC project, which provided only limited exposure to elements designed to enhance self efficacy, demonstrated enhanced m emory control beliefs, but no memory self efficacy gains, following the training program (Hastings & West, 2009). Thus, changes in memory self efficacy may require greater group interaction to provide more opportunities for vicarious experience and verbal persuasion. This evidence suggests that cognitive interventions may promote enhanced self regulation, particularly self evaluative beliefs Do self regulatory factors in turn, predict greater cognitive performance within the context of interventions? In the Senior Odyssey program (Stine Morrow et al., 2014), an inductive reasoning intervention, higher initial levels of memory self efficacy predicted greater gains in inductive reasoning (Payne et al., 2012). That is, trainees with more confidence in their memory spent more time on training activities and demonstrated greater training related improvements. Similarly, study was explained by baseline me mory self efficacy, as well as level of education and subjective health, according to discriminant analysis (Bagwell and West, 2008). Further training related change in memory self efficacy was a direct predictor of memory gains from the intervention as a ssessed using latent growth curve modeling (West & Hastings, 2011). Importantly, active trainees demonstrated greater gains in memory self efficacy than did inactive trainees (Bagwell & West, 2008). These findings underscore the
38 importance of enhanced se lf evaluative beliefs in promoting training gains in memory performance. Metamemory Research evidence suggests that cognitive interventions may enhance metamemory and that cognitive interventions which promote metamemory may be effective in promoting train ing related memory gains and transfer effects (Hertzog & Hultsch, 2000). This knowledge can be general (e.g., how memory functions; differences between memory loss in dementi a and in healthy aging) or personal (e.g., General knowledge about memory strategies, i dentifying the most effective personal strategy for a specific task situation, y adapting information processing to task demands are metamemory processes involved in cognitive self regulation (Hertzog & Dunlosky, 2011). A ging may impact the effectiveness of strategy selection : older adults are less likely than younger adults to spon taneously employ memory strategies, and they may overly rely on external aids or familiar mnemonics (McDaniel & Bugg, 2012). Therefore research paradigms relevant to metamemory and cognitive training have focused on selection and application of memory str ategies a s well as self monitoring techniques (Berry et al., 2010). The focus here will be on strategy selection. Increased strateg ic knowledge and more effective strategy application is often assumed to underlie performance benefits from training (West et al. 2000). Limited data examines this assumption, due in part to the challenges of directly assessing strategy use as when using think aloud procedures and related validity concerns (West &
39 Strickland Hughes, 2015). One way investigators might assess effectiveness of strategy use is by providing explicit instruction to use a specific strategy, and then assuming case, memory performance following strategy instr uctions becomes a proxy of the Subjective evaluation of s trategy selection can be measured using self report questionnaires of various formats (e.g., open ended, multiple choice, checklist) administered after memory testing (West & Strickland Hughes, 2015). A review of 12 interventions that included subjective evaluations of strategy selection computed standardized pre post gains in strategy use between trainees and control participants. The effect size for st rategic behavior gain for trainees was 1.86 (Gross & Rebok, 2011). The review suggests that strategy improvement was greater for 97% of trainees, compared to control participants, in these studies. The EMC project (Bagwell & West, 2008; West et al., 2008) was one of the interventions included in this analysis. Strategy selection was assessed using self report surveys in the EMC project (West et al., 2008). Immediately following list, name, and story recall tasks, participants indicated which strategies they had used by selecting from a list of commonly used strategies, with an open ended option to list any idiosyncratic techniques they had employed. These checklists were completed during pretest and posttest assessments immediately following training and one month later. Over time, trainees and control participants had a tendency to spontaneously increase the number of strategies used. More importantly, trainees, but not control participants, also transitioned to more effective use of the strategies over time Therefore, the EMC project demonstrated that
40 strateg ies may be increased and optimized via participation in a multifactorial intervention program. Furthermore the methodology employed in EMC to assess strateg ies required participants to be self aware of strategy usage that is, to be aware of strategy selection and implementation as part of their self regulatory approach to task success Strategy selection, at least for list recall tasks, may also be evaluated by examination of clustering techniques. A benefit of this approach is that it does not require participants to self report use of strategies after testing, which requires high levels of metamemory and task monitoring (Gross & Rebok, 2011). The key to measuring clusteri ng is to examine word recall order. That is, the order in which participants recall list items might be evaluated for items recalled in order of the study list (serial clustering, suggesting use of association strategies) or items recalled in categorical g roupings (semantic clustering, suggesting use of categorization), or (3) consistent use of a particular recall order across trials (subjective clustering, suggesting use of consistent or idiosyncratic strategies). Use of serial, semantic, and subjective cl ustering techniques by memory trainees and control participants were analyzed in the ACTIVE trial (Gross & Rebok, 2011). Trainees immediately increased in their use of each clustering technique, and these gains were maintained over five years. Further, thi s training effect was present regardless of various participant demographic characteristics, such as age and years of education. This evidence suggests that older adults may be more likely to successfully and frequently employ memory strategies as a conseq uence of training. In turn, effective strategy selection may directly enhance memory performance. It is possible for trainees
41 to learn how to monitor themselves and then adjust attention and revise strategy use according to judgments of their own learning or modify their approach based on the particular needs of a specific task situation (Dunlosky et al., 2007; Hertzog & Dunlosky, 2011). If this is not done, memory strategies learned in the laboratory cannot easily be applied to everyday life ( M cdaniel & Bu gg, 2012). In the EMC program, there was a tendency for participants to use simple r strategies, and few trainees indicated using all the steps of the more complex strategies (West et al., 2008). Therefore, even when strategy use increases, strategies might not always be applied in the most effective manner to promote memory on a particular task For example, compared to the overall training effect for strategic behavior reported by Gross and Rebok (2011), the magnitude of training effect for memory performa nce was considerably smaller. Unlocking the maximum potential of benefit from cognitive training requires participants to generalize strategic knowledge that is, to take strategies learned in one context and apply them more broadly To this end, the limit ed research evidence is very hopeful. For example, older adults learned two mnemonics (i.e., sentence generation and interactive imagery) and practiced them to recall lists and paired associates (Cavallini, Dunlosky, Bottiroli, Hertzog, & Vecchi, 2010). Tr ainees successfully generalized use of these mnemonics to a novel, unpracticed text memory task when simply instructed that the strategies might be helpful for other memory tasks. In contrast, performance on the novel task did not improve for trainees who did not receive this instruction. Via training, individuals may better understand which mnemonics are most effective for themselves or may more effectively pair appropriate strategies to
42 specific situations (Hertzog & Dunlosky, 2012; Hertzog & Jopp, 2010). In this project, strategy selection was a key target variable for change during training.
43 CHAPTER 4 THE PRESENT STUDY Over five decades of research on episodic memory training confirm that older adults may benefit from these programs: trainees, compared to control groups, demonstrate improved performance on trained tasks immediately following training (Gross et al. 2012). Consistent with a self regulatory perspective, proposed mechanisms of training related memory gains include superior self evaluative b eliefs, such as memory self efficacy and personal locus of control for memory (Payne et al., 2012; West & Hastings, 2011), and enhanced metamemory, such as effective strategy select ion and implementation (Bailey et al. 2010; Gross & R ebok, 2011; Hertzog e t al., 2009 ). T hese potentially c ritical self regulatory factors were the focus of this research. A n established, award winning memory training program the Everyday Memory Clinic ( EMC ) project (West et al., 2008) was used as the foundation for this research t o enhance the potential for important lessons to be learned from this work. The EMC project was a multifactorial memory training program for adults age 50 or older. The design of the EMC project incorporated several elements to enhance memory sel f efficacy across four known influential sources (i.e., mastery experience, vicarious experience, verbal persuasion, and physiological and affective states; B andura, 1997; Smith & West, 2006 ). Compared to waitlist control participants, EMC trainees demonst rated pretest posttest gains in name recall, story recall, memory self efficacy, memory control beliefs and strategy use. Latent growth curve modeling (West & Hastings, 2011) suggested memory self memory per formance and in enhancing memory: level of memory self efficacy predicted story recall and name recall performance, and gains in memory self efficacy predicted
44 gains in story recall. Thus, the EMC project was highly successful and highlighted the value of a focus on self regulation in strategy training. However, the precise value added from the focus on self efficacy is unknown, because EMC combined an emphasis on self regulatory beliefs with multifactorial memory training. That is, the intervention focused on strategy training plus beliefs and did not separately examine the impact of the se component elements from strategy training per se The present research Everyday Memory Clinic Revised (EMC R ) was designed to address this limitation and extend previous findings by evaluating the precise impa ct of beliefs focused training. U ltimately, this project aimed to maximize the impact of memory training for older adults. Specific A ims The global purpose of the EMC R research wa s to determine the impact of a relatively brief one week version of a memory intervention that emphasizes str ategy training plus self regulation, in comparison to a control group and a traditional approach focused on strategy training by itself The research also examined v alue added for the self regulation condition, in comparison to the traditional approach Va lue added wa s operationalized by greater change in primary outcomes, and by better transfer to untrained memory tasks. The primary outcomes include d name recall and self regulatory factors memory self efficacy, memory control beliefs and strategy selection. Aim 1. Effectiveness of Abbreviated Training Dosage The first aim of the present research was to determine whether a one week multifactorial training program focused on strategies for name recall is
45 effective (i.e., yields improved primary outco mes immediately following training in comparison to an inactive waitlist control group). While successful, the original EMC project was somewhat exacting and relatively lengthy: training covering multiple strategies for list, story, and name recall lasted over five weeks and included homework and readings outside of class meetings (West et al., 2008 ). EMC R aimed to extend EMC project results by examining whether a briefer training dosage could produce meaningful gains E vidence suggests that relatively br ief memory training programs can enhance performance on specifically trained memory tasks (Dunlosky et al. 2003; Gross et al., 2012; Rebok & Balcerak, 1989 ; West, 1989 ). Further, meta analyses on memory training suggest that multifactorial programs, with activities such as relaxation or attention training, may be generally more successful in enhancing memory than strategy only programs. Also, programs that train multiple strategies may be more effective than programs that train single strategies (Gross et al., 2012; Verhaeghen et al., 1992). The present research determined whether an abbreviate d one week training program can be successful in improving memory and enhancing beliefs. In EMC R multiple strategies were taught but for name recall only (not for list or story recall), and training included several different types of in class and at home exercises. Thus, EMC R wa s multifactorial, but ha d a relatively short duration (one two hour train ing session plus approximately 2 hours of homework ). To examine Aim 1, pretest and posttest scores for the primary outcomes were compared to evaluate the impact of training against the impact of repeated testing in the control group. Aim 2. Impact of Beliefs Focused Training The second aim of EMC R was to deter mine whether a beliefs focused strategy training condition compared with a matched strategy training
46 condition without the focus on beliefs yielded greater improvement for primary outcomes, from pretest to posttest This research aimed to define the benefi on enhancing self regulation. Considering evidence from the EMC project and other research, scholars recommend a focus on enhancing self regulation as part of memory training (Berry et al. 2010; Dunlosky et al., 200 7; Hertzog et al., 2009 ; Rebok & Balcerak, 1989; West & Strickland Hughes, 2015). Unfortunately, the precise benefit from a self regulatory focus in memory training is unknown. Limited previous research, such as the EMC project, has compared self regulatio n focused training to inactive control groups (Bailey et al., 2010; West et al., 2008) or to an alternate program focused on health behaviors without mnemonic instruction (McDougall et al., 2010). Other research has examined the relationship between traini ng related performance gains and self regulatory factors, such as positive self evaluative beliefs, in programs that were not specifically designed to enhanc e self regulation (Jaeggi, Buschkuehl, Shah, & Jonides 2014; Payne et al., 2012). No research to d ate has compared an EMC style training program, specifically designed to enhance self regulatory factors, with a traditional, strategy only program matched in frequency, duration, and strategy content. Comparison of these matched groups in EMC R helped det ermine the exact value added from a training methodology designed to enhance self regulatory factors in terms of changes in memory, strategy selection, and memory related beliefs Aim 3. Achieving Near Transfer The third aim of EMC R was to determine whether beliefs focused memory training promotes near transfer to untrained memory task s.
47 Enhanced self regulation may also serve as a key to maximizing the practical impact of training. As mentioned, memory training programs evidence success in improving performance on the trained tasks immediately following training. While some intervention techniques, such as extensive practice of processing resources, demonstrate modest success in promoting transfer effects (Jaeggi et al., 2008), the memory strategy tra ining research rarely reports successful transfer, even to tasks very similar to trained abilities (McDaniel & Bugg, 2012; West & Strickland Hughes, 2015). However, enhancing strategy knowledge (a type of metamemory) via mere instruction promoted successfu l transfer of benefit (Cavallini et al., 2010), and trainees with a growth mindset, similar to a greater sense of personal control, have exhibited transfer effects in working memory training (Jaeggi et al., 2014). Thus, enhanced self regulation may facilit ate transfer of training benefits ( Hertzog & Dunlosky, 2012; Jaeggi et al., 2014; Stine Morrow & Basak, 2011; Strickland Hughes & West, 201 7 ). The present research evaluated whether a b eliefs focused program heightened the practical impact of memory traini ng by promoting near transfer. Successful near transfer was tested via comparison of pretest posttest performance change on two untrained associative memory tasks across trainees in the strategy plus beliefs and strategy only conditions and participants in a wait list control group. Experimental Overview Native English speaking, community dwelling men and women aged over 50 years old were recrui ted to participate in the study. Additional eligibility criteria included freedom from cognitive impairment (per pe rformance on a telephone assessment) and sensory functioning (hearing and vision) adequate for completing telephone and in person assessments. Research participants were randomly assigned in blocks to one of
48 three experimental conditions: (1) a waitlist control group ( CT ), (2) a traditional strategy only training group ( SO), (3) or a strategy plus beliefs training group (SB; the target training group). The two training conditions were matched in terms of strategy content and training frequency and duratio n and strategy content. Both training conditions focused on strategies for name recall. All training occurred in a single week and took approximately five hours, including one 2 hour in person training followed by three hours of recommended self study in a workbook. However, t he strategy plus beliefs raining condition ( SB) but not the traditional strategy only condition (SO), incorporated several elements designed to enhance self regulation (i.e., memory self efficacy, memory control beliefs strategy sele ction) in the in person sess ion and in the workbook Training impact was evaluated using assessment of primary outcomes and near transfer effects. The primary outcomes were memory performance on the trained name recall task and self regulatory beliefs, spe cifically memory self efficacy memory control beliefs, and strategy selection. Based on self regulatory theories (Hertzog et al., 2009 ), effective strategy selection was operationalized as use of a greater number of strategies and use of the most effectiv e trained strategy Transfer effects were evaluated by comparing the three experimental c onditions on transfer outcomes including visual associative memory and verbal associative memory, with both immediate and delayed tests. Within person comparisons cha rted training benefit to primary and transfer outcomes the week before training ( P retest) and immediately following training ( P osttest). Participant eligibility, baseline ability factors, and other individual factors were
49 assessed in a phone interview befo re any other assessment or training activities. These factors were included to evaluate whether random assignment by block led to comparable groups at pretest. Hypotheses Aim 1 Hypotheses The firs t aim of the present research wa s to examine the effectiveness of an abbreviated dosage of strategy training for name recall, by comparing all trainees (SO plus SB) to t he waitlist control group (CT). No differences in the primary outcomes were expected between the trained and control groups at pretest due to random assignment Interaction effects were expected for each of the primary outcomes. Trainees were expected to demonstrate greater pre post improvements in each primary outcome (memory and self regulatory factors) than participants randomly assign ed to the c ontrol (CT) condition. Both a main effect and an interaction effect were expected for name recall performance. D ue to practice effects, both groups were expected to perform better on the name recall test at posttest than at posttest ( Hypothesis 1a ) However, the trainees were expected to demonstrate greater performance improvement on the name recall test than the control group ( Hypothesis 1 b ). An interaction effect was also expected for memory self efficacy: compared with their pretest reports, t he trainees were expected to report higher levels of memory self efficacy at posttest where as the CT group was expected to report decreased memory self eff i cacy at posttest, given that memory self efficacy can decline simply from taking memory assessments ( Hypothesis 1 c ). There were no clear predictions for memory control for trainees because there is insufficient study of this variable in past research. However, the CT group was expected to report
50 no pre post change in memory control beliefs due to repeat ed testing ( Hypothesis 1 d ). For number of strategies used, similar to name recall performance, a main effect was expected to show that all participant groups used more strategies at posttest than at pretest ( Hypothesis 1e ) However, the trainees were expected to report using a greater number of strategies than the CT group at posttest ( Hypothesis 1 f ). Further, as assessed imm ediately following training at p osttest, a greater proportion of trainees than c ontrol participants were expected to report using the most effective training strategy ( Hypothesis 1 g ). Thus, results were expected to confirm that a short term training program can be effective in improving memory and enhancing beliefs, even with brief training program. Aim 2 Hypotheses The seco nd aim of the present research was to determine the precise value added to a memory training program from using a training approach design ed to enhance self regulation, as opposed to a traditional, strategy only approach. With random assignment by block to training conditions, both groups were expected to have comparable levels of these variables at pretest. Considering the preponderance of research evidence implying success of episodic memory training programs, a main effect was expected for pre post impro vement in name recall performance ( Hypothesis 2a ) Further, a qualifying interaction of condition and test was expected SB trainees we re expected to demonstrate greater recal l improvements than SO trainees, due to an emphasis on self regulation ( Hypothesis 2 b ). A different pattern of change than that e xpected for memory performance was expected for memory self efficacy, which has been shown to decline simply from taking
51 assessments (West et al., 2008; West et al., 2003) and to improve from social but not individual, EMC style memory training (Hastings & West, 2009). An interaction of condition and test was predicted At posttest compared to p retest, the SB gr oup wa s expected to report i ncreased memory self efficacy and the SO group was expected t o report maintained memory self efficacy ( Hypothesis 2 c ). An interaction of test and condition was also expected for memory control beliefs. The SB trainees were expected to report pre post increases in memory control ( Hypothesis 2 d ). It wa s not clear whe ther there would be any changes in memory control beliefs for the SO group There is little research concerning whether standard strategy training results in any gains in memory control beliefs. Thus, there we re no specific predictions for degree of change in control for SO trainees. A main effect of condition was expected for number of strategies used: both trained groups were expected to benefit from practice effects by using more strategies at posttest ( Hypo thesis 2e ) However, a n interaction of condition and test was expected for t he number of strategies used G iven the self regulatory elements in the strategy plus belief training condition, SB trainees were expected to demonstrate a greater increase in numb er of strategies used than the SO trainees ( Hypothesis 2 f ). In addition, at posttest, a condition main effect for strategy selection was expected A greater proportion of trainees in the SB condition were expected to use the most effective trained strategy than those in the SO co ndition ( Hypothesis 2 g ) Aim 3 Hypotheses The third aim of the present research f ocused on the practical impact of memory training, that is, whether benefits to training transfer to untrained tasks as evidenced by
52 near transfer to associative memory tasks comparable to the name face association task that was trained. The transfer measures were similar in that they assess ed associative memory, but they differ ed in modality and difficulty. The object location task (delayed mat ching of objects to locations) was primarily visual, and the occupation name task (immediate and delayed matching of names to occupations and delayed recognition of correct pairs ) was primarily verbal. Although no data compare these tests, based on past results for these individual measures it was anticipated that they would vary from easy to difficult in this order: delayed location recall delayed occupation name pair recognition, immediate occupation name recall delayed occupation name recall It is possible, then, that the relative gain from transfer would vary across these outcomes, but there was insufficient evidence to make a specific prediction about such variation. These different outcomes were included so that the project may inform whether any kind of transfer occur r ed given general failure to evidence near transfer in past memory strategy training. We anticipate d that SB would be the only group to exhibit transfer, showing greater pre post gains than the other two groups ( SO and CT), who will not dif fer from each other, in object location delayed recall ( Hypothesis 3a ), occupation name delayed pair recognition ( Hypothesis 3b ), occupation name immediate recall ( Hypothesis 3 c ), and occupation name delayed recall ( Hypothesis 3 d ). If confirmed, these hypotheses would suggest that enhancing self regulation is an effective avenue to facilitate transfer after memory strategy training.
53 C HAPTER 5 METHODS Participants Community dwelling m iddle aged and older m en and women of any race or ethnicity were recruited from the Gainesville, Florida, area using newspaper advertisements, existing participant pools, lifelong learning programs, and word of mouth Participants paid a fully refundable deposit of $25 to receive the training materials, which were valued at $45. The deposit was waived for two participants who did not bring the deposit with them to the training class. If participants decided not to do the training and interviews but returned the training materials, the full deposit was returned. For compensation, when participants completed the research part of the training program (pretest and posttest interviews), their full deposit was returned and they could keep the training materials. Participant eligibility was screened using a n initial interview conducted by telephone. Inclusion criteria for the study were vision, hearing, and English skills adequate to complete the interviews and training, freedom from cognitive impairment, no reported major stroke or head injury in the previous year, more than eight years of education, and age of at least 50 years. Twenty one individuals completed the phone interview but were ineligible for the study due to performance on the Telephone Interview for Cognitive Status (TICS; Brandt, Spencer, & Folstein, 1998) scoring below the recommended cut off score of 31, which was suggestive of cognitive impairment. An additional 23 individuals completed the telephone interview but did not participate in any other part of the research program, due to scheduling issu es ( n = 10), illness ( n = 9), death in the family ( n = 2), and unspecified reasons ( n = 2).
54 Eligible participants scheduled for the complete research program included 130 middle aged men and women. Eight of th o se participants were excluded for inability to follow interview instructions during the in person assessments. The final sample included 122 participants of whom 117 completed the entire research program. On e participant did not complete the first in person assessment due to scheduling issues. F our participants did not complete the second in person assessment, withdrawing after completion of training due to unspecified reasons ( n = 2), health issues ( n = 1), and lack of interest ( n = 1) Two of these participants were assigned to the strategy onl y training group, and the other two were assigned to the strategy plus beliefs training group, suggesting that the attrition was not due to condition assignment. P articipants who dropped out of the study after the pretest and training session were similar to those who completed the posttest interview. Indep en dent samples t tests compared mean scores of these two groups for age, self rated vision, hearing, physical health, and mental health years of education, number of prescription medications taken regula rly, baseline cognitive performance (Telephone Interview for Cognitive Status, Backward Digit Span, RAVLT immediate and delayed list recall), and all primary was not significant (all p 05) for the comparison of any of these measures across both groups, suggesting that the assumption of homogeneity of variance was tenable. Mean scores for participants who dropped out of the study following the pretest interview were comparable to those fo r participants who completed the posttest interview, except that those who dropped out had lower scores on the Telephone Interview for Cognitive Status ( M diff = 2.44, SE = 1.06) t (120) = 2.30, p = .023, rated their correct vision as
55 poor er ( M diff = 1.54, SE = 0.67), t (119) = 2.27, p = .025, and reported using fewer strategies on the face name association test ( M diff = 10.17 SE = 5.06), t (119) = 2.01 p = .047 suggesting that ability factors may have influenced this attrition The m iddle aged and older p articipants were 51 to 93 years old ( M = 73.24 years, SD = 8.31 years) and half of the participants were under 75 years old Most participants identified as female ( 78.7%) and as White or Caucasian (91.9%). P articipants were highly educated an d healthy. total years of education ranged from 11 to 27, with a mean of 17.33 years ( SD = 2.84 years). Self reported health was assessed with the SF 12 (Ware, Kosinski, & Keller, 1995). Norm based standardized scores for physical health and mental health are computed to have means of 50 and standard deviations of 10 i n the general U.S. population. I n the present study, scores for physical health ranged from 18.83 to 64.05, with a mean of 49.04 and standard deviation of 9.35, and scores for me ntal health ranged from 26.11 to 65.02, with a mean of 54.68 and a standard deviation of 6.58. Participants reported taking an average of 2.66 pres cription medications regularly. About one third (30.8%) of participants reported taking one or fewer prescription medications regularly and only 8.3% reported taking five or more prescription medications regularly. Only 16 participants (13.3%) reported being hospitalized in the previous year. Recruitment of this sample size was inform ed by a priori power analyses conducted in G*Power (version 3.0.10; Faul, Erdfelder, Lang, & Buchner, 2007), and the final sample size was consistent with those used in previous research (Hastings & West, 2009; West et al., 2008). Although effect sizes may decline over time (Open Science Collaboration, 2015) m ost relevant effect sizes from the EMC project were
56 large (West et al., 2008). The p ower derived from this sample size was sufficient to detect medium large effect sizes in the analyses (e.g., one way ANOVAs with three groups; repeated measures ANOVAs with two measurements and three groups, with between repetition correlations for outcomes estimated as r = .40 .80, based on memory self efficacy and name recall data from Strickland Hughes et al., 2016 ) Research Design EMC R was a memory strategy training intervention for middle age d and older adults that utilized a randomized control trial design. Based on availability, with no information provided about the type of training they would receive, researc h participants were scheduled to attend particular testing and training sessions. Research participants were randomly assigned by session to one of three training conditions: a waitlist control group (CT; n = 38) a traditional strategy only traini ng group ( SO; n = 46) or the target strategy plus beliefs training group, designed to enhance self regulatory factors as well as memory (SB; n = 38) A ssignment was based on session enrollment to ensure adequate enrollment size in each training class ( n = 12 20, based on the Everyday Memory Clinic) and led to relatively similar numbers across conditions. The design of the research was a 2 (time: pretest, p osttest) x 3 (training condition: CT, SO, SB ) mixed model design with time as a within subject vari able and training condition as a between subject variable. Research participants, procedures, and all measures central to the primary research aims and hypotheses are described in detail in the following sections.
57 Assessment Procedure Completion of the res earch program (e.g., training and in person assessments) for each participant (i.e., pretest, training, posttest) lasted three weeks An overview of the research program is outlined in Table 5 1 Before starting the research program, all participants complete d an initial telephone interview, which served as an eligibility screening and was used to assess demographics and other relevant control measures, such as cognitive capacity. During the first week o f the program, all participants complete d the pretest assessment of primary outcomes (i.e., name recall, memory self efficacy, memory control beliefs, and strategy use). During the second week of the research program, the training groups ( SB and SO) comple ted training, with or without embedded elements to enhance self regulation, while the waitlist control group (CT) did not participate in any meeting, homework, or other activity Posttest assessments, matched in type and difficulty to the pretest assessmen ts, were administered in the third week of the training program. The phone interview initiate d the verbal informed consent process and last ed approximately 30 45 minutes. Baseline measures were collected during phone interviews using established measuremen t tools, known to be reliable and valid for phone administration to older adults. These measures were used to confirm participant eligibility and to assess the similarity of experimental groups produced by the random assignment by block to training conditi ons. Research assistants audio recorded the telephone interviews after obtaining participant consent. All participants consented to the audio recording of their telephone interview. The recordings were reviewed by Strickland Hughes and supervisors for qua lity control purposes.
58 Procedures for the pretest and p osttest assessments were nearly identical, although different versions of memory tasks were administered in a counterbalanced order for participants across all conditions. An abbreviated version of instructions for survey measures and strategy checklists were initially provided at the posttest interview, but the full instructions from the pretest procedures were administered if participants had questions. Pretest and p osttest assessments were adminis tered in small group testing sessions ( n = 1 13) in meeting rooms at the University of Florida campus or public access rooms in the local area Of the 45 different testing sessions (pretest and posttest), the mean class size was 5.56 participants ( SD = 3 .63 participants ), and five sessions had only one participant scheduled. Testing sessions include d participants from across all three training conditions to ensure that the self regulatory advantages from the SB group did not carry over into being tested class. training conditions. Carla M. Strickland Hughes attended 12 of 21 (57%) posttest sessions and Robin L. West attended 1 of 21 (5%) of th e posttest sessions. Strickland Hughes and West were not always blind to participant condition because they during the interviews, which had more limited interaction with participants than the The p retest assessment began with written informed consent, followed by administration of paper and pencil tasks. Measures assessed at the pretest and p osttest assessments examined primary outcomes (i.e ., name recall and self regulatory factors), transfer outcomes (i.e., object location and occupation name memory ), and a
59 few exploratory measures. The Pretest and Posttest assessments lasted approximately 1.5 2 hours. The complete list of measures, with citations, administered during the phone interview and pretest/posttest assessments appears in Table 5 2 and each measure is described in detail next. Measures and Materials Background and control measures Background and control measures were assessed during the phone interview. Background measures include d medication and supplement use, demographic information (e.g., years of education, gender, date of birth) and ratings of English language ability and se nsory function. Additional measures included a screening for cognitive function ( Telephone Interview for Cognitive Status; TICS), a test of episodic memory ( Rey Auditory Verbal Learning Test; RAVLT), a test of working memory (Backward Digit Span), and surv eys of self reported health (SF 12), general personal control beliefs (P erceived Mastery memory ability (General Memory Evaluation; GME). Each of these measures are known to be reliable and valid if administered over the p hone to older adult populations Telephone Interview for Cognitive Status (TICS). The Telephone Interview for Cognitive Status (Brandt et al., 1988) is a telephone screening for cognitive impairment designed for older adults. The TICS includes 11 questions, with a maximum score of 41 points. Higher scores indicate better cognitive performance. Example items include asking participants to identify the complete date, their current location, and the first and last names of the American president and vice president. Participants are also asked to count backwards by sevens from 100 five times, to provide antonyms for common words, name specific items, repeat phrases, and tap on the phone five times. The TICS
60 is highly sensitive and specific, is strongly correlated to performance on the widely used Mini Mental State Examination, r = .94, and has strong test retest reliability, r = .97. Following the guid elines tested by Brandt et al. (1988), participants eligible for the present research scored greater than 30 out of 41, suggesting freedom from cognitive impairment. Rey Auditory Verbal Learning Test (RAVLT). The Rey Auditory Verbal Learning Test (Lachman, Agrigoroaei, Tun, & Weaver, 2014; Lezak, 1995) assesses episodic verbal memory. RAVLT includes immediate and delayed free recall of a list of 15 commonplace nouns, such as flower and school During encoding, research assistants read the 15 words, with app roximately 1 second pause between each word. Participants were given 90 seconds immediately following encoding to recall as many of the words as they could remember. Following a n approximate ten minute delay participants were given 60 seconds to recall th e words again. Scores were calculated as the number of words correctly recalled ranging from 0 to 15 Higher scores represent better episodic memory performance. Number of intrusions provided were also counted. Backward Digit Span test. The Backward Digit Span test, adapted from the WAIS III for use in the Brief Test of Adult Cognition by Telephone (Lachman et al., 2014; Wechsler, 1997), assesses working memory span. Research assistants read string s of digits at a rate of one second per digit. Participants were asked to repeat the digits in the reverse order. Digit span increase d from a length of two digits to a length of eight digits If participants missed the first string at one level (length of string ), a second item at that level was administered. The task was discontinued if a participant incorrectly answered both items at a single level. The backward digit span s core was
61 calculated as the longest digit span correctly repeated backwards at least once with a maximum score of 8. Higher scores repres ent greater working memory capacity. Short Form Health Survey (SF 12). Self reported physical health and mental health were assessed using the Short Form Health Survey (SF 12; Ware, Kosinski, & Keller, 1996). The SF 12 includes twelve items that use variou s response scales, ranging from dichotomous response options of Yes / No to 6 point Likert scales. One item assesses overall general health ( In general, would you say your health is with responses from 1 = excellent to 5 = poor ), and the other items asses s the extent to which persons are limited as a result of physical or emotional problems (e.g., During the past four weeks, how much did pain interfere with your normal work, including both work outside the home and housework? with 5 point Likert scale res ponse options from Not at all to Extremely ) Summary scores f or the physical component scale (PCS) and the mental component scale (MCS) were calculated following proce dures outlined by Ware and colleagues ( 1995 ). These norm based standardized scores are continuous and calculated to have a mean of 50 and a standard deviation of 10 in the U.S. adult population. H igher values represent better physical or mental health. The inte rnal consistency of the SF 12 was adequate ( = .81 ), consistent with p revious research (Strickland Hughes et al., 2016) although the items were selected to represent several dimensions of physical and mental health and are thus heterogeneous. The SF 12 has good relative validity when compared to a longer form of the health s urvey and other health criteria and measures (Ware et al., 1996). Perceived Mastery survey. The Perceived Mastery scale (Lachman & Weaver, provided a
62 baseline score representing a global analog to the domain specific personal locus of control for memory of interest in this research. The present research used the 5 item version of the Perceived Mastery scale adapted from Pearlin and Schooler (1978) by Lachman and Weaver (1998) and used in t he Health and Retirement Study (HRS; http://hrsonline.isr.umich.edu ; Hauser & Willis, 2005 ) a large scale longitudinal study Example items include: When I really want to do something, I usually find a way to succeed at it and What happens to me in the future mostly depends on me Responses were made using a 6 point Likert scale ranging from 1 = strongly disagree to 6 = strongly agree A perceived mastery score was computed as the mean of the responses to thes e five items (range: 1 6) Higher scores represent greater sense of perceived mastery. The scale had adequate internal consistency 78, although this reliability was not as strong as that reported by the 2006, 2008, and 2010 w aves of HRS .90) Internal consistency would not have been higher if any of the items were removed. General Memory Evaluation survey (GME). The General Memory Evaluation survey (West, Dark Freudeman, & Bagwell, 2009) assesses global beliefs about memory, comparable to the specific memory self efficacy measure of interest in this study and thus represents a baseline score for memory self efficacy The three items in the GME concern evaluation of recent memory performan memory to same aged peers, and overall satisfaction with recent memory performance. An example item is How satisfied are you with your recent memory performance? Participants rate three items using a 7 point Likert scale (e.g., 1 = very unsatisfied to 7 = very satisfied ). An index GME score was calculated by averaging responses to the three
63 items (range: 1 to 7) Higher scores indicate greater perceived general memory ability. This score had good internal consistency reliability, Cro previous research (Strickland Hughes et al., 2016) Primary o ut come m easures Primary outcome measures were assessed at the pretest and posttest. The first primary outcome measure was name recall performance on the face name association memory task which was the episodic memory task for w hich strategies were trained. The other primary outcome measures were the self regulatory factors that were i dentified as related to episodic memory and expected to benefit from tra ining: memory self efficacy, memory control beliefs and str ategy use. Name recall. Memory test procedures for name recall followed EMC procedures and use d the same materials (West et al., 2008). Participants completed two levels of the name recall task. A t each level, testing of name recall immediately follow ed encoding of name face pairs and participants were aware of the upcoming assessment. At encoding, participants were instructed to study names and faces. T he to be remembered stimuli were names print ed in all capital letters underneath color portraits of ethnically di verse men and women of all ages as described in greater detail in West et al. (2008) Name face pairs were printed four per page and presented in a notebook. An example page is included in Figure 5 1 At testing, greyscale versions of the same portraits were presented in a different order and without names P articipants were asked to write r face in their testing packet. Level 1 of the task include d 12 name face pairs Participants were allotted one minute for encodi ng and four minutes for recall. Level 2 of the task was a more c hallenging version of Level 1. At Level 2, t he Level 1 name face pairs were presented
64 again, in a new order, with an additi onal 12 name face pairs. For the 24 name face pairs at Level 2 participants received five minutes for encoding and five minutes for recall. This tiered structure of the memory assessment should have allowed for the maximal influence of self regulatory fac tors by allowing individuals to invest additional effort over time on the second trial, if willing to do so Two different sets of name face pairs that were of comparable difficulty based on past research (West et al., 2008) and that had similar balances o f race, gender, and age of the faces were used. Set assignment was counte rbalanced by testing occasion (pretest, p osttest) and training condition. Each participant was assigned a different set of name face pairs at pretest and p osttest The two sets were of comparable difficultly. Name recall performance did not vary according to set of name face pairs at pretest for Level 1, t (119) = 1.34, p = .182, or for Level 2, t (119) = 0.89, p = .375, or at posttest for Level 1, t (115) = .56 p = 574 or at Level 2, t (115) = 0.032 p = 975 Name recall scores were calculated as the percent of correct ly written names at each level. The primary outcome measure of interest was name recall performance on the Level 2 version of the task, given that t his level is more difficult and should provide participants greater opportunities to self regulate their performance. Scoring criteria were strict; for a response to be scored as correct, the exact name needed to be written, and nicknames, similar names, o r partial names were not accepted. For example, neither Bob nor Robbie would be scored as correct for Robert. Spelling errors were acceptable if the written name would have been pronounced the same way as the correct name (e.g., Robbert for Rob ert; Carlie for Carly)
65 Memory Self Efficacy Questionnaire (MSEQ 4). Memory self efficacy strength memory task. Memory self efficacy strength was assessed using the Memory Self Efficacy Questionnaire 4 (MSEQ 4; West, Thorn, & Bagwell, 2003). MSEQ 4 includes four subscales wherein respondents indicate how confident they are that they can remember names, object locations, shopping list items, and sto ries. Confidence ratings range from 0 = I cannot do it to 100 = 100% sure I can do it in ten percent increments. Each memory task is rated for five levels of difficulty, for instance, ranging from remembering all 18 shopping list items for a sick friend to recalling only 2 of those items. Participant responses to all 20 items were averaged for a single score of memory self efficacy strength (range: 0 100). Higher scores indicate a greater level of confidence ks. The MSEQ 4 had very good in ternal consistency reliability (pretest ). Personal locus of control for memory. memory control beliefs were a ssessed using the Locus subscale of the Metamemory in Adulthood Questionnaire (MIA; Dixon, Hultsch, & Hertzog, 1988). The Locus subscale of the MIA consists of nine items, each a nswered using a 5 point Likert scale. Example items include I have little control over my memory ability and If I were to work on my memory I could improve it A memory control beliefs index score was calculated by averaging all responses. A higher score ( range: 1 5) indicate d greater sense of personal control over memory ability The Locus subscale of MIA had adequate internal consistency reliabil ity (pretest ), which was consistent with previous
66 literature ( i .71 .78 across four studies; Dixon et al., 1988). Name recall strategy use. Strategy use was assessed using a ch ecklist of 1 6 strategies for name recall. Using procedures from the EMC project (West et al., 2008) this checklist was administered following Level 2 name recall testing. Participants were instructed to check each memory strategy that they used on the task and to write any idiosyncratic strategies use d that were not on the list. Example strategies include I tried to pick out prominent features and I repeated names over and over to myself Participants also indicate d the one or two strategies that they used the most. The strategy checklist is reproduced in Appendix A. One dependent variable was number of strategies used. Greater strategy use indicates greater self regulation because it indicates that participants flexibly employed multiple strategies. Number of strategies used w as calculated as the perce ntage of the 17 listed strategies that were checked were tabulated only if the accompanying written note described a valid strategy not already included in the checklist. As in the EMC project, data scored from the strategy checklist also enable d examination of whether participants across the training conditions employed the most effective strategies and/or strategies that were taught in the training program. The most effective strategies were identified by comparing name recall performance between participants who did and did not use each strategy. The proportion of trainees in each condition who indicated using the most effective strategy as their most frequent strategy used w as another indicator of strategy use.
67 Transfer outcome measures T ransfer outcome measures were administered at the pretest and p osttest assessments. Transfer outcome measures were performance on the two untrained, associative memory tasks object location visual association and occupation name verbal association. As these measures were adapted from similar paradigms reported in the literature, test timing and difficulty for this project were fine tuned with pilot testing. Two forms of each task wer e developed and counterbalanced across testing occasion and training condition so that participants received a different set of items at pretest and posttest. Posthoc paired t tests confirmed that the different sets were of comparable difficulty; performan ce did not vary between the two sets at pretest or posttest for any of the object location or occupation name memory tests (all p s > .05) Object location visual association. Procedures for the object location visual association memory were adapted from t hose outlined in West, Welch, and Knabb (2002) and Welch (1998). Location recall was assessed using a matrix array task wherein participants placed 24 pictures of objects (e.g., hat, scissors) in a 3 4 picture array representing 12 rooms in a house and then later recalled their own placements. A complete list of the objects, example pictures of objects, and an example room array are included in Appendix B. Participants were given five minutes to place t he objects in the array and to study their placements Objects could be placed in any room with no more than two objects in each room. Following a delay of approximately 40 minutes, participants were allotted five minutes to reconstruct their earlier place ments of the objects in the arrays. Research assistants recorded participant placements at encoding and testing using digital cameras. Object location recall was calculated as the
68 percentage of objects placed in the same room at encoding and testing. Hi gh e r percentage score s reflect ed better location memory. Occupation name verbal association. The occupation name verbal association task assessed immediate recall delayed recall, and delayed recognition memory for names paired with occupations. Occupation name stimuli were adapted from Craigie and Hanley (1997) and James (20 04) Two different sets of 30 occupation name pairs were used. Within each set, half of the occupations were concrete and able to be visualized. Eight of the concrete surnames were occupations ( e.g., FARMER), and seven of the concrete surnames were other things that could be visualized (e.g., STONE). The two sets were matched by length of names (letters and syllables) and frequency of names ( per data published by the United Sta tes Census Bureau from the 2000 Census), as well as general familiarity with the occupations and how easy the concrete names were to visualize, per ratings from an independent sample ( N = 14). A complete list of the occupation name stimuli is included in A ppendix C. Encoding and testing procedures were adapte d from Cavallini et al. (2010), and participant instructions for the task are included in Appendix C. During encoding, participants were given six minutes to study 30 pairs of names and occupations. To be remembe red stimuli were presented in a cards with the occupation name pairs printed in all capital letters. Occupations were centered on the left half of the cards, and surnames were centered on the right half of the cards. Two different orders of the occupations were randomly generated and used for the immediate recall and delayed recall testing, otherwise the administration and scoring procedures were identical. Participants we re given four minutes to write the
69 name that was paired with each occupation on a list of the 30 occupations that they had studied. Occupation name recall scores were calculated as the percentage of names correctly recalled at each test using strict codin g criteria, as described in the name recall section. Immediate recall testing occurred right after encoding, and d elayed recall testing occurred approximately 40 minutes following encodi ng. Occupation name delay recognition testing occurred approximately 7 5 minutes following encoding. During recognition testing, participants were shown a list of 30 occupation name pairs All the names and occupations were previously studied, but only half of the names were matched correctly to the occupations. Participants were given three minutes to circle delayed recognition scores was calculated as the percentage of correct hits (pair correct and indicated correct) and correct rejectio ns (pair incorrect and indicated incorrect) out of the 30 trials. Higher percentage correct scores on the immediate recall, delayed recall, and delayed recognition tests indicate better memory performance. Exploratory measures The following exploratory me asures were administered at the p retest and p osttest assessments. These measures were selected because research suggests that they are related to memory performance and/or self regulation and may benefit from training or moderate training effects, but they were not central to any of the study aims or hypotheses. Memory anxiety. Memory anxiety, or perceptions of the relationship between anxiety and memory performance, was be assessed with the 14 item Anxiety subscale of the MIA (Dixon et al., 1988). Example items include I find it harder to remember and I feel jittery if I have to introduce someone I just met As with
70 the Locus subscale of the MIA, item responses were made using a 5 point Likert scale, and an index score was calculated by averaging all responses. A higher score (range: 1 5) indicated greater memory related anxiety. The Anxiety subscale had good internal consisten c = .87 = .89 ), consistent with previous literatur e ( .83 .87 across four studies; Dixon et al., 1988). Mindful Attention Awareness Scale (MAAS). The State Mindfulness subscale of the Mindful Attention Awareness Scale (Brown & Ryan, 2003) represents the extent to which individuals pay att ention to the pr esent moment. Mindful attention may benefit from the strategy training to t he extent that name recall strategies aim to focus attention. Participants rated f ive items using a 7 point Likert scale, anchored at 0 = not at all, 3 = somewhat and 6 = very much. Example items include I find myself doing things without paying attention and present. An index score was calculated by averaging the fi ve responses (range: 0 6). H igher scores will represent greater attentiveness. MAAS had acceptable internal consistency reliability ( = .85 = .84 ), although somewhat lower than previous research with a larger sample across adulthood = 92; Brown & Ryan, 2003 ). Subjective Age Identity. Subjective age reflects how old individuals feel. Subjective age may be important for the present research, given the preponderance of negative aging stereotypes about memory and their effects on performan ce (Chasteen, Kang, & Remedios, 2011; Hummert, 2011). Participants responded to five subjective age identity questions that were adapted from previous research (Kastenbaum, Derbin,
71 Sabatini, & Artt, 1972) and administered in Strickl and Hughes et al. (2016 ) by indicating the ages that they feel. They reported their felt age ( At this moment, how old do you feel? ) and desired age ( If you could choose your age, how old would you want to be? ), among others. Proportional subjective age identity scores were calcul ated using the Chronological Age) / Chronological Age 100. Proportional subjective age identity younger or older than his or her chronological age. For example, if a woman aged 75 years old reported an average subjective age of 50 years, her proportional age identity score would be 33.33, indicating that she feels 33% younger than she is. This meas ure had strong test retest reliability, r = .97 in past research (Strickland Hughes et al., 2016), and adequate internal consistency reliability .82 ; .82 ). H ealth and memory engagement information. A survey assessing health and memory engagement non ability factors that might affect memory performance was administered. This survey was created for the present research, and t he pretest and posttest versions are included in Appendix D. The health related items, such as whether the participant has been hospitalized recently and a list of current medications taken, were collected for post hoc examination in case participants had difficultly following instructions or seemed confused. Memory engagement items included four frequency ratings of memory activities (e.g., exercise my memory with games or activities not on a computer, a = never to i = daily ) and reports of participation i n other formal memory training programs. An index score was calculated by averaging responses to the four
72 frequency ratings (range: 1 9), with higher scores indicating a greater frequency of engagement in memory related activities. This scale had weak in ternal reliability consistency 64 .71 ), which was unsurprising given that each of the items assessed the frequency of discrete memory related activities. Training Procedures Training was completed for the r elevant groups (SB and SO) in the second week of the research program. Training duration and strategy content was the same for both training groups. All trainees initially complete d a two hour class with in person instruction on strategies for name recall led by Carla M. Strickland Hughes or Robin L. West. Classes were video recorded to establish quality control and instructor consistency for the program At the conclusion of the tr aining class, participants received training materials including a written transcript of the training class, additional readings on strategies for name recall, and practice exercises. Using this workbook, trainees were assigned approximately two to three h ours of continued self study as homework. The homework activities were assigned to be completed in the same week as the training class and before the following interview. Trainees were asked to document their training related activities on a diary form. Th e homework assignment sheets for both training conditions are included in Appendix E In addition, trainees were provided lab contact information (phone number and email address) and invited to contact the research team with any questions about the trainin g materials. No participants, in either training condition, contacted the lab with questions about the homework materials.
73 Training materials for the two groups (SO and SB) did not vary in strategy instruction or practice exercises. That is, t he strategy r elated content of the readings, examples, and exercises in the initial strategy lessons and workbook were the same for both training groups. All trainees, but not the Control group, learned association and visualization strategi es for names paired with fac es, such as the image name match method as well as active observation attentional techniques The image name match method involves several steps, with the goal of visualizing an object that represents the r Bill) next to an exaggerated version of a prominent or distinctive facial feature. The instructor explained each strategy, modeling its use and providing several examples, then trainees practiced each technique during in class activities. For example, e ach trainee came up with an association for his or her own name and shared with the group, and trainees practiced using all steps of the image name match method with several novel face name pairs. The printed training materials included a written transcrip t of the training class, additional rea dings on name recall strategies, e.g., the How to Remember Names chapter and sev eral practice exercises for recognition and recall of faces, names, and name face pairs. Self regulatory elements in training Although the strategies taught did not differ by training group, the t raining format varied according to whether the approach was designed to enhance self regulation. That is, the SB strategy lessons and workbook included e lements to enhance self regulation, maintaining the EMC style approach To the extent possible, these training elements were absent for the SO training group to determine the precise benefit from a training focus on enhancing self regulation (Aim 2). Some elements, such as social
74 support provided by peers, may occur naturally in group training; the instructor emphasize d these features for the SB group, but not the SO group The target elements used to enhance self regulation were self efficacy theory; they target four influential sources of self efficacy: enactive mastery, vicarious experience, verbal persuasion, and physiologic and affective states (Bandura, 1997; Smith & West, 2006 ). T he purpose of the SB trai ning was presented as one f ocused on personal improvement, rather than getting 100% correct ( reduces anxiety when individuals have personal co ntrol over the effort required), and practice exercises in the SB version of the workbook were in order of increas ing challenge with simpler exercises first (to gain mastery before trying more difficult tasks). SB trainees received positive feedback on use of the trained strategies as observed by the instructor in class al for memory change) and heard reframing of negative comments from the instructor (anxiety reduction). Throughout the SB program, instructors emphasize d that memory change is possible at any age. In contrast, t he SO training condition did not emphasize any self regulatory elements: the focus of the training was presented as obtaining high performance practice exercises were presented in random order, and instructors ignore d rather than reframe d negative comments Instructors did not provi de positive feedback, nor did they emphasize the p otential for growth at any age. A complete listing of the elements to enhance self regulation included in each condition is detailed in Table 5 3 and specific examples of these elements in the training mat erials are reproduced in Table 5 4 Expert raters coded the written training materials for elements that might enhance or diminish self efficacy. Written materials that varied between the SO and SB
75 training groups were coded. These materials included a tra nscription of the training session, instructions, and assigned readings. Expert coders tallied incidences of specific categories of elements that should enhance self efficacy, such as anecdotes or personal stories of memory success, mention of research evi dence and that training can be effective, and emphasis on individual differences in ability and that individuals are in control over their own strategy use (see Table 5 3 ). Absolute agreement amongst the five expert raters who coded all written training ma terials (i.e., instructions, assigned readings that differed between SB and SO conditions, transcript of the training sessions) was high; the intraclass correlation coefficient for their ratings was 0.95. The SB written training materials ( M = 140.40, SD = 14.15) had significantly more incidences of the elements to enhance self efficacy across all rated categories than the SO written training materials ( M = 95.20, SD = 9.65), t (4) = 12.77, p < .001. Paired samples t tests also suggested that the SB version of the materials had a greater number of each of the different types of elements to enhance self efficacy, as compared to the SO version of the materials.
76 Table 5 1 Week by week overv iew of as sessment and training procedure Week Groups Agenda SR 0 All Phone interview: Eligibility, demographics, and other control measures n/a 1 All Pretest assessment: Primary and transfer outcomes n/a 2 SB 2 hour group training session, followed by ~ 2 hours of self study in workbook SO 2 hour group training session, followed by ~ 2 hours of self study in workbook Control No meeting, homework, or other activity 3 All Posttest assessment: Primary and transfer outcomes n/a *These groups may have include d some self regulatory elements, just because of their social nature, but no self regulatory elements were supported or emphasized by the instructor or the materials. Note: SR = Inclusion of elements to enhance self control.
77 Table 5 2 Overview of measures with references Measure Source Phone Interview Verbal Informed Consent Telephone Interview for Cognitive Status (TICS) Brandt et al., 1988 Rey Auditory Verbal Learning Test (RAVLT) Lachman et al., 2014; Lezak, 1995 Backward Digit Span from WAIS III Lachman et al., 2014; Wechsler, 1997 Short Form Health Survey (SF 12) Ware et al., 1996; Ware et al., 1998 Perceived Mastery Lachman & Weaver, 1998 General Memory Evaluation (GME) West et al., 2009 Surveys of medication/supplement use, ratings of sensory function, demographic information Pretest/Posttest Assessments Written Informed Consent Name Recall (Levels 1 and 2) West et al., 2008 Memory Self Efficacy Questionnaire (MSEQ 4) West et al., 2003 Locus of Control for Memory (Metamemory in Adulthood Questionnaire subscale) Dixon et al., 1998 Occupation name Verbal Association Craigie & Hanley, 1997; James, 2004 Object Location Visual Association Welch, 1998; West et al., 2002 Strategy Use Checklists (for name recall, occupation name association, and object location association) West et al., 2008 Memory Anxiety (Metamemory in Adulthood Questionnaire subscale) Dixon et al., 1998 Mindf ul Attention Awareness Scale (MAAS) Brown & Ryan, 2003 Subjective Age Identity Kastenbaum et al., 1972; Strickland Hughes et al., 2016 Surveys of health and memory engagement Created for the present research Pretest only
78 Table 5 3 List of training elements to enhance self regulation included in each condition Self efficacy elements Control SO SB Enactive mastery Each skill practiced repeatedly In class success remembering names of class members S uccess with strategy practice Easier strategies first Self set goals for practice and achievement Vicarious experience Trainer gives examples, models each strategy Emphasis on learning from each other in class Encourage personal stories provided by group members Whole group practices strategy together Verbal persuasion Group provides social support Emphasize what "you already know" Positive feedback from train er in session Readings emphasize potential at any age Research presented on strategy effectiveness and learning potential of seniors Trainer reframes all negative comments to emphasize potential Physiologic and affective states (reducing anxiety) At home training materials to allow self pacing or sufficient time for learning Readings reviewed in class to help trainees understand Class discussion shows that others have similar problems* Emphasis on process/learning, not on 100 % score Focus on potential in readings and in class sessions Self set goals, emphasize personal decision making about what to learn Trainer emphasizes that everyone has memory failures While this element may occur naturally in group training, trainer will emphasize or encourage this in SB condition but not in the SO condition.
79 Table 5 4 Examples of self regulatory elements in training materials SB Version SO Version 1. Workbook I ntroductory Paragraph 1. Workbook Introductory Paragraph We recognize that each person begins on a different level. We all start with different skills and different ability levels. Each person has different memory experiences. By beginning with fairly simple techniques, we hope to ensure that all of you understand and benefit from this lesson, even if you have very little background in memory. The Everyday Memory Training program is not about being an expert improving, so that you will be able to remember more after the program is done. Memory is complicated. We have been working on memory research for many years. So we are going t o try to teach you specific techniques that might help you to improve your memory performance. Because there are many different techniques, you will learn more than one technique in class. Some of these techniques will be hard and other memory methods will be easy. And you will be asked to do practice exercises with these techniques. The lesson in class today and the readings and practice exercises are intended to help you to raise your memory scores. 2. Workbook Closing Paragraph 2. Workbook Closing Paragraph In this memory training program, you have learned how to use different memory techniques. We have discussed association, imagery, the image name match method and active observation. Research evidence tells us that each of these techniques can su pport your memory in general and make it more likely that you will remember names. The homework section of this notebook starts on the next page. The homework includes readings and memory activities that will give you an opportunity to practice and master the memory techniques that you learned today. In this memory training program, you have learned how to use different memory techniques. We have discussed association, imagery, the image name match method and active observation. Each of these techniques can support your memory in general and make it possible to remember names. The homework section of this notebook starts on the next page. The homework includes readings and memory activities that will improve your memory scores. 3. Excerpt from General Instr uctions for Practice Exercises 3. Excerpt from General Instructions for Practice Exercises You have an activity log with this packet so that you can keep track of your exercise time. The more you practice, the better you will do on these tasks. It is up t o you. With additional practice, you are more likely to master the strategies you learned in class. You have an activity log with this packet so that you can keep track of your exercise time. The more you practice, the better you will do on these tasks.
80 Figure 5 1 Example page of to be remembered name face pairs used in the name recall task
81 CHAPTER 6 RESULTS The research design employed a 2 (time: pretest, posttest) 3 (condition: CT, SO, SB) mixed model design that allowed for examination of research aims via separate repeated measures analyses of variance (ANOVAs) and chi square tests. Primary outcomes were continuous variables (name recall performance, memory self efficacy, personal locus of control for memory, number of strategies used) and one categorical variable (use of most effective strategy). All transfer outcomes were continuous variables (object location recall, occupation name immediate recall, occupation name delayed recall, occupation name delayed recognition). Reported effect sizes are partial eta 2 ) values from SPSS version 24 (IBM Corp., 2016). Preliminary Results: Baseline Data by Condition Preliminary univariate ANOVAs were conducted to en sure that the three training groups were comparable at baseline At baseline, the training groups did not differ in years of education, subjective ratings of English language ability (overall speaking, writing, and comprehension), vision, or hearing, perfo rmance on tests of episodic and working memory (RAVLT immediate and delayed recall; backward digit span), self reported physical health and mental health, perceived mastery (global control), or General Memory Evaluation (global memory self efficacy). Means and standard deviations by training condition appear in Table 6 1 2 tests confirmed that random assignment to training condition was independent from whether participants self identified as female or white, p s > .05. Although the SB condition appeared relatively disadvantaged across several baseline measures (e.g., fewer years of education, worse self reported physical health),
82 the only significant condition difference was for age at pretest, F (2,121) = 3.50, p = .033. Follow up pairwise compar isons confirmed that the SB participants were older than the SO participants ( M diff = 4.57, SE = 1.78, p = .035) but not the CT participants ( M diff = 3.56, SE = 1.87, p = .176), and the SO and CT participants did not differ in age ( M diff = 1.01, SE = 1.78, p = 1.00). To address this condition difference for participant age, all analyses were conducted with and without age included as a covariate. Because the pattern of results was the same, the analyses are reported without age as a covariate. Aim 1 Results: Effectiveness of Abbreviated Training Dosage The first aim of the present research was to test whether brief strategy training for name recall is effective. Training effectiveness was operationalized as greater improvement for trained participants compared to the waitlist control group, for the primary outcomes. Separate 2 time (within: pretest, posttest) 2 condition (between: control, trained) repeated measures ANOVAs were conducted for each of the continuous primary outcomes (name recall, mem ory self efficacy, memory control, number of strategies used). Means and standard deviations are reported in Table 6 2. F approximations are reported for all repeated measures ANOVAs with M test of homogeneity of covariance ( p < .001). All pairwise post hoc comparisons were conducted with Bonferroni corrected analyses, and p values are reported where significant. Alpha level of 0.05 was used as the significance criteria. A 2 test was conducted for the categorical primary outcome variable (whether most effective strategy was used).
83 Training Effects for Name Recall Performance Level 2 name recall performance was examined because this more difficult level of the t est was expected to benefit from training, due to its nested design. As expected ( Hypothesis 1a ), the main effect of time was significant, F (1,115) = 3.61, p 2 = .03. Name recall performance at posttest was greater than performance at pretest acro ss both trainees and control participants. Qualifying this main effect, the interaction between time and condition was significant, F (1,115) = 4.32, p 2 = .04. Post hoc pairwise comparisons indicated that performance between trainees and control p articipants did not differ at pretest ( M diff = 0.85, SE = 4.83, p = .861) or at posttest ( M diff = 6.81, SE = 4.77, p = .156). However, confirming Hypothesis 1b trainees performed better at posttest than they did at pretest ( M diff = 7.33, SE = 2.10, p < 001), whereas control participants did not ( M diff = 0.33, SE = 3.03, p = .914). The interaction effect is depicted in Figure 6 1. Training Effects for Memory Self Efficacy For memory self efficacy, the main effect of time was unexpectedly significant, F (1,115) = 27.62, p 2 = .19. Memory self efficacy was greater at posttest than at pretest across both trainees and control participants ( M diff = 5.81, SE = 1.12, p < .001). This main effect was qualified by a significant interaction between time an d condition, F (1,115) = 7.51, p 2 = .06, consistent with Hypothesis 1c Post hoc pairwise comparisons indicated that memory self efficacy between trainees and control participants did not differ at pretest ( M diff = 3.85, SE = 3.31, p = .391) or a t posttest ( M diff = 3.31, SE = 3.24, p = .310). However, trainees reported greater memory self efficacy at posttest than they did at pretest ( M diff = 8.99, SE = 1.28, p < .001), whereas control
84 participants, who were expected to report declined pretest to posttest memory self efficacy, reported no change ( M diff = 2.83, SE = 1.85, p = .129), as shown in Figure 6 2. Training Effects for Memory Control Beliefs A 2 time (within: pretest, posttest) 2 condition (between: control, trained) repeated measures ANO VA was conducted for memory control beliefs. Unexpectedly, the main effect of time was significant, F (1,115) = 7.10, p 2 = .06. Memory control beliefs were higher at posttest than at pretest across both trainees and control participants ( M diff = 0 .11, SE = 0.04, p = .009). The interaction between time and condition was not significant, F (1,115) = 2.29, p 2 = .02. Post hoc pairwise comparisons indicated that memory control beliefs did not differ between trainees and control participants at pretest ( M diff = 0.03, SE = 0.10, p = .773) or at posttest ( M diff = .15, SE = 0.10, p = .131). As expected ( Hypothesis 1d ), control participants reported no pre post change in memory control beliefs due to repeated testing ( M diff = .05, SE = 0.07, p = .485). However, trainees reported pre post gains in memory control beliefs ( M diff = .17, SE = 0.05, p < .001). Training Effects for Strategy Use Number of strategies used A 2 time (within: pretest, posttest) 2 condition (between: control, trained) repeat ed measures ANOVA was conducted for number of name recall strategies used. Consistent with expectations ( Hypothesis 1e ), the main effect of time was significant, F (1,115) = 26.72, p 2 = .19. Both trainees and control participants used a greater nu mber of name recall strategies at posttest than pretest ( M diff = 5.76, SE = 1.11, p < .001). Further, the interaction between time and condition was significant, F (1,115) = 5.70, p 2 = .05. Post hoc pairwise comparisons indicated that
85 strategy use did not differ between trainees and control participants at pretest ( M diff = 0.64, SE = 1.98, p = .746). However, trainees used a greater number of strategies than control participants at posttest ( M diff = 4.67, SE = 2.33, p = .047), confirming Hypothesi s 1f Control participants reported no pre post change in strategy use ( M diff = 3.10, SE = 1.83, p = .093), but trainees reported pre post gains in strategy use ( M diff = 8.41, SE = 1.27, p < .001). The significant interaction is depicted in Figure 6 3. Effective strategy use Effective strategies were operationalized as name recall strategies that were associated with superior name recall performance. Independent samples t tests were conducted to test whether mean name recall performance (percent correct at Level 2) varied as a function of using a particular strategy (the complete strategy checklist is included in Appendix A). The two tailed significance levels of = .05 were used, although a priori expectations were directional (use of a given strategy w as expected to be related to better performance than not using the strategy). Chi square tests were also used to examine whether use of the most effective strategies as one of the one or two most frequently used strategies (that is, as the preferred strate gy ) varied between trainees and the control group. Tests were conducted separately for pretest and posttest data. Identification of effective strategies. One strategy was identified as effective at both pretest and posttest. That most effective strategy w as I associated the name with the name of someone else I know (i.e., familiar name strategy). Participants who reported using the familiar name strategy, compared to those who did not, performed better on the Level 2 name recall test at pretest (47.93% use d; M diff = 14.91, SE = 4.27),
86 t (119) = 3.49 p < .001, and at posttest (54.24% used; M diff = 16.45, SE = 4.21), t (116) = 3.90 p < .001. Another strategy seemed to be effective overall. Participants who reported using the strategy I tried to think of a me aningful association for the name (i.e., associate name strategy) performed better than those who did not at pretest (57.02% used; M diff = 10.76, SE = 4.41), t (119) = 2.44 p = .016, and use of the associate name strategy was marginally related to better p erformance at posttest (65.25% used; M diff = 8.35, SE = 4.62), t (116) = 1.81 p = .073. Thus, the second most effective strategy was use of a meaningful association for the name. Two other strategies were effective at pretest or posttest, but not both testing occasions. Participants who reported using the strategy I associated the face with the face of someone else I know (i.e., familiar face strategy) performed comparably to those who did not at pretest (37.19% used; M diff = 5.15, SE = 4.61), t (119) = 1.12, p = .266, yet had superior performance at posttest (50.00% used; M diff = 13.77, SE = 4.28), t (116) = 3.22, p = .002. Participants who reported using the strategy I covered the faces, looked away and tested myself on the names (i.e., self test strateg y) performed better than those who did not at pretest (14.05% used; M diff = 17.74, SE = 6.24), t (119) = 2.84, p = .005, but this benefit was not evidenced at posttest (28.82% used; M diff = 5.59, SE = 4.90), t (116) = 1.16, p = .248. The two strategies that were generally effective (familiar name and associate name), as well as the other strategy that was effective at posttest (familiar face), related to use of associations. Forming associations was a focus of the training class and represented a relatively s imple strategic approach, compared to the more complex
87 strategies taught in class. Therefore, these three strategies were also examined together as a set of effective strategies. This definition of effective strategy use was more lenient than the other two because the familiar name and associate name strategies were included in this identified effective strategy set Examination of effective strategy use by training condition. Chi square tests of independence were conducted to evaluate whether frequency of effective strategy use varied between trainees and control participants at pretest and posttest. Most effective strategy use was defined as preferred use of either (a) the effective familiar name strategy, (b) the effective associate name strategy, or (c) the set of three effective association strategies. Use of this tiered classification system permitted a nuanced examination of training effects for this self regulatory factor. At pretest, trainees and the control participants did not differ in their frequ ency of 2 (1) = 0.32, p = .570, of the associate 2 (1) = 0.09, p 2 (1) = 1.36, p = .244. Trainees and control participants also did not differ in t heir frequency of using the familiar name strategy as a preferred strategy at posttest, 2 (1) = 0.14, p = .714 This finding suggested that the training did not affect on whether participants relied on making associations between the names and the names of people whom they knew. However, this particular strategy was not emphasized during training. A greater proportion of trainees, compared to control participants, reported 2 (1) = 4.88, p = .027. The odds ratio suggested that the odds of preferred use of the associate name strategy at posttest were 2.80 times higher if participants received training than if they did not
88 receive training. Further, there was a significant association between training and 2 (1) = 5.08, p = .024. Based on the odds ratio, the odds of participants preferring one of the effective association strategies at posttest were 2.48 times higher if th ey received training than if they did not receive training. These findings, showing greater use of association strategies by trainees, provide partial support for Hypothesis 1g. Results of the chi square tests are summarized in Table 6 3. Aim 2 Results: Im pact of Beliefs Focused Training The second aim of the proposed research was to determine the precise value added from training designed to enhance self regulation (strategy plus beliefs; SB), as opposed to a traditional, strategy only approach (strategy o nly; SO). Repeated measures analyses were used to compare training effectiveness (i.e., greater improvements for the primary outcomes) between the two training approaches. Separate 2 time (within: pretest, posttest) 2 condition (between: SO, SB) repeated measures ANOVAs were conducted for each of the continuous primary outcomes (name recall, memory self efficacy, memory control, number of strategies used). All pairwise post hoc comparisons were conducted with Bonferroni corrected analyses, and p values ar e reported where significant. Alpha level of 0.05 was used as the significance criteria. Means and standard deviations are reported in Table 6 4. A 2 test was conducted for the categorical primary outcome variable ( whether most effective strateg y was used). Beliefs Focused Training Effects for Name Recall As expected ( Hypothesis 2a ), the main effect of time on Level 2 name recall performance was significant, F (1,77) = 12.74, p 2 = .14. Across both training
89 conditions, trainees demonstrat ed improvement at posttest, compared to their pretest performance ( M diff = 7.37, SE = 2.07, p = .001). Although a significant interaction effect was expected ( Hypothesis 2b ) results suggested that the pre post improvement in name recall performance was no t greater for SB trainees than for SO trainees, F (1,77) = 0.03, p 2 < .001. Name recall performance did not differ between SB and SO trainees at pretest ( M diff = 8.28, SE = 5.70, p = .150) or at posttest ( M diff = 7.55, SE = 5.53, p = .176). Perf ormance improved at posttest, compared to pretest, for both SB trainees ( M diff = 7.74, SE = 3.08, p = .014) and SO trainees ( M diff = 7.01, SE = 2.75, p = .013). Beliefs Focused Training Effects for Memory Self Efficacy For memory self efficacy strength, the main effect of time was significant, F (1,77) = 50.15, p 2 = .39, and the interaction between time and condition was not, F (1,77) = 1.04, p 2 = .01. Memory self efficacy strength was greater at postte st than pretest ( M diff = 8.84., SE = 1.25, p < .001) across both SO and SB trainees. SB trainees reported levels of memory self efficacy strength that were similar to SO trainees at pretest ( M diff = 2.45, SE = 3.85, p = .526) and at posttest ( M diff = 5.0 0, SE = 3.74, p = .184). Further, memory self efficacy strength was higher at posttest than pretest for both SB trainees, ( M diff = 7.57, SE = 1.86, p < .001) and SO trainees ( M diff = 10.12, SE = 1.66, p < .001). Consistent with expectations ( Hypothesis 2c ) the SB trainees reported pre post increases in memory self efficacy strength, but, contrary to expectations, SO trainees also reported increased, rather than maintained, memory self efficacy strength.
90 Beliefs Focused Training Effects for Memory Control A 2 (time: pretest, posttest ) 2 (condition: SO, SB) repeated measures ANOVA with personal control beliefs for memory as the dependent variable was conducted to test Hypothesis 2d The main effect of time was significant, F (1,77) = 15.12, p 2 = .16, but the interaction between time and condition was not, F (1,77) = 1.18, p = .280, 2 = .02. Memory control beliefs were greater at posttest than pretest ( M diff = 0.18, SE = 0.05, p < .001) across both SO and SB trainees. SB trainees reported levels of memory control beliefs that were similar to SO trainees at pretest ( M diff = 1.29, SE = 0.11, p = .262) and at posttest ( M diff = 0.03, SE = 0.12, p = .798). Consistent with expectations ( Hypothesis 2d ), memory control was higher at posttest than prete st for SB trainees ( M diiff = 0.22, SE = 0.07, p = .001). Memory control was also higher at posttest than pretest for SO trainees ( M diiff = 0.13, SE = 0.06, p = .039). Beliefs Focused Training Effects for Strategy Use Number of strategies used A 2 time (wit hin: pretest, posttest) 2 condition (between: SO, SB) repeated measures ANOVA was conducted for number of name recall strategies used. The main effect of time was significant, F (1,77) = 39.26, p 2 = .34. A greater number of name recall strategie s were used at posttest than pretest across all trainees ( M diiff = 8.24, SE = 1.31, p < .001), as expected ( Hypothesis 2e ). However, the interaction between time and condition was not significant, F (1,77) = 1.33, p 2 = .02, contrary to expectations ( Hypothesis 2f ). Number of strategies used was similar for SB and SO trainees at pretest ( M diiff = 1.96, SE = 2.24, p = .384) and at posttest ( M diiff = 1.08, SE = 2.63, p = .683), and strategy use increased from pretest to posttest for both
91 SB trainees ( M diiff = 6.72, SE = 1.96, p = .001), and SO trainees ( M diiff = 9.76, SE = 1.75, p < .001). Effective strategy use As described in the Aim 1 Results section, effective strategies were defined as those related to superior name recall performance. The most effective strategy was familiar name, the second most effective strategy was associate name, and a set of three effective association strategies was also identified. Chi square tests of independence were conducted to evaluate w hether frequency of effective strategy use varied between SB and SO trainees at pretest and posttest. These results showed no differences between SB and SO trainees as summarized in Table 6 5, and provided no support for Hypothesis 2g Aim 3 Results: Nea r Transfer Effects The third aim of the proposed research was to determine whether the beliefs focused training program would promote near transfer, as evidenced by pre post performance gains on transfer tasks, or untrained associative memory tasks. Given the limited reports of near transfer effects reported for past memory training programs, four different transfer tasks were administered to permit a broad and nuanced examination of whether any near transfer might occur. The transfer tasks were object loca tion recall (OBJ), occupation name immediate recall (ON Imd), occupation name delayed recall (ON Delay), and occupation name delayed recognition (ON Recog). Preliminary analyses were conducted to evaluate the correlations between the transfer tasks and to compare their difficulty. Hypotheses 3a 3d were tested with separate repeated measures 2 time (within: pretest, posttest) 3 condition (between: CT, SO, SB) ANOVAs that compared performance on the four transfer tasks to determine whether
92 the beliefs fo cused training program promoted near transfer, compared to the other groups. All pairwise post hoc comparisons were conducted with Bonferroni corrected analyses, and p values are reported where significant. Alpha level of 0.05 was used as the significance criteria. Means and standard deviations are reported in Table 6 6. Preliminary Analyses for Transfer Outcomes The four transfer outcomes were expected to be related to each other and to the targeted name recall task, but to represent distinct memory perfor mance variables. Bivariate correlations between each of the transfer outcomes at pretest and at posttest were calculated to assess the magnitude and direction of the relationships between these variables and with the targeted face name association task. P r correlation coefficients are reported in Table 6 7. All correlation coefficients were significant ( p < .001). Relationships among the transfer outcomes ranged from r = .34 to r = .97, with the weakest relationships between the object location te st and the occupation name tests and the strongest relationships between the immediate and delayed occupation name recall tests. Performance on the occupation name recall, occupation name recognition, and object location recall tasks seemed to be distinct. Relationships between level 2 name recall performance and the transfer outcomes ranged from r = .46 to r = .62. At pretest and at posttest, the magnitude of the relationships between name recall and the occupation name immediate and delayed recall tasks w ere greater than those of the name recall task with the occupation name delayed recognition task or the object location delayed recall task. These findings suggest that the transfer tasks overall were sufficiently unrelated to the name recall task, and tha t training related improvements on the occupation name recall tasks might
93 represent more proximal transfer than improvements on the occupation name delayed recognition task or the object location recall task. A within subjects ANOVA was conducted to compa re mean performance for all participants on the four transfer outcomes (test type: OBJ, ON Imd, ON Delay, ON Recog) at pretest to assess the relative difficulty level of these associative memory tasks. The effect of test type was significant, F (3,357) = 23 0.22, p 2 = .66. Pairwise comparisons indicated that performance on each of the transfer outcomes differed from performance on the others at pretest, all p s < .001, suggesting that these tasks varied in difficulty. The measures varied in difficul ty following the expected pattern: by order of performance from low to high, tasks were occupation name delayed recall, occupation name immediate recall, occupation delayed recognition, and object location delayed recall, suggesting that the visual object location task was easier than the verbal occupation name task overall. For the occupation name task, recall was more challenging than recognition, and recall was more challenging after a delay that allowed for additional interference. Mean scores and 95% c onfidence intervals are illustrated in Figure 6 4. Near Transfer Effect for Object Location Recall Near transfer to object location recall was tested with a 2 time (within: pretest, posttest) 3 condition (between: CT, SO, SB) repeated measures ANOVA. Ne ither the main effect of time, F (1,109) = 0.05, p 2 < .001, nor the interaction between test and condition, F (2,109) = 0.57, p 2 = .01, were significant. Performance on the object location delayed recall test did not differ across the three conditions at pretest or at posttest, and posttest performance did not differ from pretest performance for any
94 of the three conditions, all p s > .05. Contrary to Hypothesis 3a no support for transfer to performance on the object location recall task was evidenced. Performance on this task did not improve as a function of training or repeated testing. Near Transfer Effect for Occupation Name Dela yed Recognition Near transfer to occupation name delayed recognition was tested with a 2 time (within: pretest, posttest) 3 condition (between: CT, SO, SB) repeated measures ANOVA. The main effect of time was significant, F (1,115) = 18.89, p 2 = .14. Overall, mean performance was greater at posttest than at pretest ( M diiff = 7.46, SE = 1.71, p < .001). Hypothesis 3b was not supported as the interaction between time and condition was not significant, F (2,115) = 2.22, p 2 = .04. Pairwise comparisons suggested that occupation name delayed recognition performance did not vary by condition at pretest or at posttest, all p s > .05, although the SB trainees trended towards worse performance at pretest than both the SO trainees ( M diiff = 12.79, SE = 5.44, p = .061) and the control group ( M diiff = 12.21, SE = 5.63, p = .097). A follow up univariate ANOVA comparing occupation name recognition performance at pretest across the three training conditions was significant, F (2,119) = 4.46, p 2 = 07. SB trainees performed worse than both SO trainees and the control group at pretest, suggesting baseline differences in the training conditions may have limited their capacity to demonstrate near transfer effects. However, pairwise comparisons that f ollowed the repeated measures ANOVA, showed that posttest occupation name recognition performance for SB trainees was improved at posttest, compared to pretest ( M diiff = 12.59, SE = 3.10, p < .001) and the same was true for SO trainees ( M diiff = 5.83, SE = 2.80, p = .039), but not for the control group ( M diiff = 3.95, SE = 3.01, p = .193). These
95 results suggest that some transfer occurred, but not just for the SB trainees as expected. Near Transfer Effect for Occupation Name Immediate Recall Near transfer to occupation name immediate recall was tested with a 2 time (within: pretest, posttest) 3 condition (between: CT, SO, SB) repeated measures ANOVA. The main effect of time was significant, F (1,115) = 15.95, p 2 = .12. Overall, mea n performance was greater at posttest than at pretest ( M diiff = 5.26, SE = 1.32, p < .001). T he interaction between time and condition was marginally significant, F (2,115) = 2.57, p 2 = .04. Performance on the occupation name immediate recall task did not differ across the three conditions at pretest or at posttest (all p s > .05), although SB trainees appeared to perform at a relative disadvantage at pretest. However, performance was better at posttest than pretest for both SB trainees ( M diiff = 6. 21, SE = 2.15, p = .005) and SO trainees ( M diiff = 8.43, SE = 2.38, p = .001), whereas the control group demonstrated no pre post change in performance ( M diiff = 1.11, SE = 2.31, p = .623). This result reflected transfer, but did not support Hypothesis 3c which predicted transfer only for the SB group Occupation name immediate recall performance by condition at pretest and posttest is illustrated in Figure 6 5. Near Transfer Effect for Occupation Name Delayed Recall Performance on the occupation name imme diate recall and delayed recall tasks were highly correlated, and the pattern of results for these two transfer outcomes was the same. Near transfer to occupation name delayed recall was tested with a 2 time (within: pretest, posttest) 3 condition (betw een: CT, SO, SB) repeated measures ANOVA. The main effect of time was significant, F (1,115) = 20.85, p 2 = .15.
96 Overall, mean performance was greater at posttest than at pretest ( M diiff = 5.77, SE = 1.26, p < .001). T he interaction between time an d condition was marginally significant, F (2,115) = 2.87, p 2 = .05. Performance on the occupation name delayed recall task did not differ across the three conditions at pretest or at posttest (all p s > .05). The SB trainees again appeared to perform with relative disadvantage at pretest, but the difference was not significant. However, performance was better at posttest than pretest for both SB trainees ( M diiff = 9.54, SE = 2.28, p < .001) and SO trai nees ( M diiff = 5.83, SE = 2.06, p = .005), whereas the control group demonstrated no pre post change in performance ( M diiff = 1.91, SE = 2.22, p = .386), reflecting transfer for both trainee groups (lack of support for Hypothesis 3d ). Occupation name delay ed recall performance by condition at pretest and postt est is illustrated in Figure 6 6
97 Table 6 1. Baseline statistics by training condition CT ( n = 38) 78.9% female 94.7% white SO ( n = 46) 76.1% female 89.1% white SB ( n = 38) 81.6% female 92.1% white Total ( N = 122) 78.7% female 91.8% white M SD M SD M SD M SD Age (years) a 72.51 8.34 71.51 7.34 76.08 8.83 73.24 8.31 Years of education 17.76 3.07 17.50 2.96 16.68 2.36 17.33 2.84 Rating of English ability (1 10) 9.32 0.93 8.91 1.31 9.03 0.97 9.07 1.11 Rating of corrected vision (1 10) 8.08 1.44 8.17 1.06 7.68 1.56 7.99 1.38 Rating of corrected hearing (1 10) 7.87 1.55 8.33 1.52 7.97 1.64 8.07 1.57 RAVLT immediate recall (0 15) 7.68 2.41 7.28 2.23 7.21 2.71 7.39 2.42 RAVLT delayed recall (0 15) 4.11 2.47 4.37 2.41 4.34 2.91 4.28 2.60 Backward digit span (2 8) 5.13 1.42 5.26 1.34 5.11 1.20 5.17 1.31 Physical health (PCS from SF 12) 50.25 8.79 49.05 10.14 47.81 8.99 49.04 9.36 Mental health (MCS from SF 12) 54.67 7.12 54.39 6.60 55.03 6.13 54.68 6.58 Perceived mastery (1 6) 5.19 0.65 5.00 0.77 4.99 0.91 5.06 0.78 General Memory Evaluation (1 7 ) 4.59 1.24 4.18 1.22 4 .04 1.24 4.26 1.24 a Significant condition difference, p < .05. CT = Waitlist control participants. SB = Strategy plus beliefs training group. SO = Strategy only training group.
98 Table 6 2. Means and standard deviations for primary outcomes for trainees and control group at pretest and posttest Control ( n = 38) Trained ( n = 79) Total ( N = 117) M SD M SD M SD Name recall ( % correct) Pretest 47.37 22.52 46.52 25.34 46.79 24.37 Posttest 47.04 23.29 53.85 24.53 51.64 24.25 Memory self efficacy (0 100) Pretest 41.96 16.45 39.11 16.95 40.03 16.77 Posttest 44.79 16.08 48.09 16.58 47.02 16.42 Memory control (1 5) Pretest 3.42 0.54 3.45 0.50 3.44 0.51 Posttest 3.47 0.47 3.62 0.52 3.57 0.51 Name recall strategy use ( % used) Pretest 26.78 10.39 26.14 9.87 26.34 10.00 Posttest 29.88 12.26 34.55 11.55 33.03 11.94 Note. Values reported are actual means, not estimated means.
99 Table 6 3. E ffective strategy use for trainees and control group at pretest and posttest Pretest Posttest Control ( n = 38 ) Trained ( n = 83 ) 2 test Control ( n = 38 ) Trained ( n = 80 ) 2 test Familiar name strategy (% preferred use ) 28.95 24.10 2 (1 ) = 0.32 p = 570 31.58 35.00 2 (1) = 0.14 p = .714 Associate name strategy (% preferred use) 31.58 28.92 2 (1) = 0.09, p = .766 18.42 38.75 2 (1) = 4.88, p = .027 E ffective association strategies (% preferred use) 63.16 51.81 2 (1) = 1.36 p = 244 50.00 71.25 2 (1) = 5.08 p = 024
100 Table 6 4. Means and standard deviations for primary outcomes for SO and SB trainees at pretest and posttest SO ( n = 44 ) SB ( n = 35 ) Total ( N = 79 ) M SD M SD M SD Name recall ( % correct) Pretest 50.19 25.59 41.90 24.61 46.52 25.34 Posttest 57.20 23.59 49.64 25.37 53.85 24.53 Memory self efficacy (0 100) Pretest 40.19 16.85 37.74 17.22 39.11 16.95 Posttest 50.31 15.85 45.31 17.27 48.09 16.58 Memory control (1 5) Pretest 3.51 0.47 3.38 0.54 3.45 0.50 Posttest 3.63 0.56 3.60 0.49 3.62 0.52 Name recall strategy use ( % used) Pretest 25.27 8.38 27.23 11.51 26.14 9.87 Posttest 35.03 10.98 33.95 12.36 34.55 11.55 Note. SB = Strategy plus beliefs training group. SO = Strategy only training group. Values reported are actual means, not estimated means.
101 Table 6 5. Effective strategy use for SO and SB trainees at pretest and posttest Pretest Posttest SO ( n = 46 ) SB ( n = 37 ) 2 test SO ( n = 44 ) SB ( n = 36 ) 2 test Familiar name strategy (% preferred use) 23.91 24.32 2 (1) = 0 00, p = 965 27.27 44.44 2 (1) = 2.57 p = .109 Associate name strategy (% preferred use) 21.74 37.84 2 (1) = 2.59 p = 108 43.18 33.33 2 (1) = 0.81 p = 368 Effective association strategies (% preferred use) 54.35 48.65 2 (1) = 0.27 p = 606 68.18 75.00 2 (1) = 0.45 p = 503 Note. SB = Strategy plus beliefs training group. SO = Strategy only training group.
102 Table 6 6. Means and standard deviations for transfer outcomes by training condition at pretest and posttest CT ( n = 38) SO ( n = 44) SB ( n = 36) Total ( N = 118) M SD M SD M SD M SD Occupation name immediate recall (%) Pretest 28.86 16.93 30.00 21.39 22.04 19.86 27.20 19.72 Posttest 30.00 19.25 36.21 20.84 30.46 22.14 32.46 20.78 Occupation name delayed recall (%) Pretest 25.00 15.02 26.06 21.74 18.61 18.91 23.45 19.04 Posttest 26.93 17.85 31.89 20.99 28.15 22.29 29.15 20.40 Occupation name delayed recognition (%) Pretest 53.51 22.15 54.09 24.45 41.30 25.98 50.00 24.71 Posttest 57.46 21.87 59.92 21.51 53.89 24.06 57.29 22.38 Object location delayed recall (%) Pretest 62.99 20.36 66.70 24.34 56.55 24.39 62.40 23.40 Posttest 63.76 19.93 64.73 21.27 58.57 25.05 62.61 22.09 Note. CT = Waitlist control participants. SB = Strategy plus beliefs training group. SO = Strategy only training group. Values are actual means, not estimated means.
103 Table 6 7. Correlations for pretest and posttest name recall and transfer outcomes Measure 1 2 3 4 5 1. Name recall (level 2) .60 .63 .51 .46 2. Occupation name immediate recall .60 .96 .74 .45 3. Occupation name delayed recall .62 .97 .75 .46 4. Occupation name delayed recognition .47 .71 .71 .52 5. Object location delayed recall .47 .41 .45 .34 Note. All coefficients are significant at p < .001. Intercorrelations for pretest are presented above the diagonal, and intercorrelations for posttest are presented below the diagonal. Each correlation was conducted using the maximum number of participants who had complete data for both measures (pr etest n = 120 122; posttest n = 114 118).
104 Figure 6 1. Name recall performance by time for trainees and control pa rticipants Error bars represent 95% confidence intervals.
105 Figure 6 2. Memory self efficacy by time for trainees and control participants Error bars represent 95% confidence intervals.
106 Figure 6 3. Name recall strategy use by time for trainees and control participants. Error bars represent 95% confidence intervals.
107 Figure 6 4. Memory performance (% correct) on transfer outcomes at pretest for all participants. Error bars represent 95% confidence intervals. OBJ = Object location delayed recall. ON Delay = Occupation name delayed recall. ON Imd = Occupation name immedi ate recall. ON Recog = Occupation name delayed recognition.
108 Figure 6 5. Occupation name immediate recall performance by condition at pretest and posttest. Error bars represent 95% confidence intervals.
1 09 Figure 6 6. Occupation name delayed recall performance by condition at pretest and posttest. Error bars represent 95% confidence intervals.
110 CHAPTER 7 DISCUSSION Everyday memory ability such as learning and remembering names is highly valued by people of all ages, yet is known to decline norma tively with increasing age, when related non ability self regulatory factors, such as self evaluative beliefs and metamemory, are also at increased risk. Collective results from cognitive interventions demonstrate that memory can be improved at any age, bu t controversy surrounds the true value of these programs in terms of the scope and duration of training related gains. A traditional and common approach to cognitive interventions for adults is memory strategy training, and limited work of this type has ex amined whether self regulatory factors might benefit from these programs or moderate other training related gains. Further, while interventions focused on intensive practice or core capacity training have demonstrated near transfer after training, perfor mance improves on untrained tasks related to the target task evidence of near transfer from strategy training programs is very rare. The present research addressed both self regulation and transfer issues. Could a short term training program, focused on strategy training, lead to changes in beliefs and strategy usage, as well as score gains? Would a short program emphasizing self regulatory change be more effective than a program focused solely on strategy training? Would near transfer be possible if both performance and self regulation changed as a function of training? Overall, this research demonstrated that a brief name recall strategy training program is effective for middle aged and older adults The training program constituted two hours of in pers on instructor led training and approximately two to three hours of
111 continued self study at home. Compared to performance at pretest and to performance of inactive waitlist control participants, training produced improved name recall memory, higher levels o f memory self efficacy, and more effective use of memory strategies. Contrary to expectations, benefits from training were similar for a beliefs focused strategy training approach, as compared to a content and duration matched training approach without th e focus on beliefs. However, benefits from both training approaches extended beyond the targeted name recall task that was trained to performance on similar, but untrained, associative memory tasks. Effectiveness of Brief Memory Training Program The first aim of this study, EMC R, was to test the effectiveness of an abbreviated memory strategy training program focused on techniques for name recall. Consistent with previous research (Gross et al., 2012; Verhaeghen & Marcoen, 1996), the training program was e ffective in terms of performance gains on the target task. The primary contribution to the cognitive intervention literature here is that EMC R was also effective in enhancing memory self efficacy and effective strategy use. The finding that EMC R enhanced self regulatory factors has several practical implications, particularly performance enhanced self evaluative beliefs and more effective strategy use that is important to p ractitioners and trainees alike. Most training programs have not directly assessed gains in self evaluative beliefs for middle aged and older adults, but among programs that did only some were successful (West & Strickland Hughes, 2015). A meta analysis b y Floyd and Scogin (1997) suggested that memory training can improve subjective beliefs about memory, although this effect was of much smaller magnitude than the benefit to memory
112 performance. Importantly, EMC R was unique from the past successful training programs because of its shorter duration and lower intensity. The origina l EMC project (West et al., 2008 memory self efficacy, but training lasted five weeks. ACTIVE trainees demonstrate d improved sense of personal control beliefs, but the ACTIVE training program lasted ten weeks (Wolinsky et al., 2009), and personal control benefits were limited to trainees in the speed of processing and reasoning training conditions, not memory strategy trainees. An 8 week long group training with 2 hour weekly meetings was effective in enhancing memory self efficacy one week and four months following training (Valentijn et al., 2005), when statistically controlling for age, gender, educational level, an d baseline scores. Older adults with memory complaints reported enhanced memory control beliefs immediately following six weeks of training with weekly 2 hour classes, but without any improvement to objective memory performance (Rapp, Brenes, & Marsh, 2002 ). Clearly, the duration of these training programs far exceeded that of EMC R, and there has been limited evidence of the benefits to beliefs from short term training. In of memory training and found that the short version was not as effective as the long one, the training was designed to last approximately 13 hours over as many days, s ignificantly longer than EMC R. Thus, the self evaluative beliefs gains in EMC R are remarkable given that this training was comparatively short in relation to other successful programs.
113 In addition to enhanced self evaluative beliefs, EMC R trainees repor ted more effective strategy use than control participants. Evidence of effective strategy use by EMC R trainees is important because strategy use is often assumed but rarely assessed (West et al., 2000; West & Strickland Hughes, 2015) and because it furthe r supports enhanced self regulation from this training program. Despite the rare assessment of strategy use in training programs, researchers often assume that memory gains are due to use of the trained strategies. Indeed, a review of strategy use across t welve interventions suggested that trainees increased strategy use over time, compared to control participants (Gross & Rebok, 2011), but much more research is needed to determine the extent to which strategy use may be more effective following training. F or example, few EMC trainees reported using all the trained strategies, and many reported only using the simpler parts of advanced strategies (West et al., 2008). This finding is consistent with research that suggests older adults may over rely on familiar strategies (Fisher, 2012). Effective strategy use represents self regulation in that it requires flexible adaptation to the task at hand, in response to continued monitoring of performance. This self regulation via effective strategy use was indicated by EMC R trainees: EMC R trainees reported using a greater number of strategies overall and a greater proportion of trainees, compared to control participants, reported preferred use of strategies that were related to superior performance. This suggests that trainees were selectively focusing their strategic effort on techniques that were advantageous. Theoretical and empirical work considers that self regulation may be key to maximizing the impact of training on memory (Hertzog & Dunlosky, 2012; West &
114 Stric kland Hughes, 2015). Self evaluative beliefs are related to greater training related improvements in cognition (Jaeggi et al., 2014; Payne et al., 2012; West et al., 2008; West & Hastings, 2011). Beliefs are also related to concurrent memory performance (Beaudoin & Desrichard, 2011), and predict memory performance up to six years later (Valentijn et al., 2006). ACTIVE memory trainees, compared to control participants, used a greater number of memory strategies up t o five years following training (Gross & Rebok, 2011), and in turn training related memory improvements were also maintained over this period (Willis et al., 2006). These benefits of training are notable, but training can only be effective if it can be com pleted. The true strength of EMC R is the possibility of broad dissemination, given its brief duration. Training that occurs over a single week, with only one 2 hour class, could easily be offered as part of existing programming at Senior Centers and lifel ong learning programs, whereas extended training would not be compatible across a wide range of settings. To this end, future work should assess the external validity of EMC R. Warranted first are assessments of the efficacy of the program when administere d by non memory experts, such as activities staff or volunteer peer leaders, with minimal training (e.g., Noice & Noice, 2013), and when samples of trainees are more diverse in terms of level of education, race and ethnic background, and gender. Additiona lly, scholars suggest that training should focus on specific memory tasks that are important to trainees (McDaniel & Bugg, 2012; Woolverton et al., 2001), and some of EMC aged and older adul ts place on name recall (Dark Freudeman et al., 2006; Strickland Hughes et al., 2016). Thus, replications of this brief training focused on other types of
115 everyday memory, such as story recall, which was improved in EMC (West et al., 2008), would logically extend the impact of this training. Beliefs Focused Versus Strategy Only Training Approaches The previous EMC project, on which this research was based, applied self s elf regulation, as well as memory performance (West et al., 2008). Because EMC research compared trainees to inactive control participants, the specific impact of this training design was unknown. Considering the modest evidence for self regulatory benefit s from relatively long training programs (discussed above), it may be that these gains were derived simply from completing memory strategy training, and not from a training approach focused on self regulation. The present research, EMC R, aimed to clarify this issue by comparing a beliefs focused training approach to a content and duration matched traditional approach focused on strategies only. Contrary to expectations, no relative benefit for the beliefs focused group was found, in terms of gains in memo ry performance or self regulation or near transfer. Potentially, training differences between the beliefs focused and strategy only conditions were comparable due to the brief time course of training and testing. First, additional time post training may be necessary to reveal the relative benefits from a beliefs focused training approach. In EMC, self evaluative beliefs were assessed one month, but not one week, following training (West et al., 2008), and latent growth curve modeling suggested that self eva luative beliefs and training related gains in these beliefs influenced memory performance one month following training (West & Hastings, 2011). Further, a primary technique for name recall taught in EMC and EMC R was the image name match technique, but thi
116 require additional time following training to maximize the potential of this strategy. Indeed, as in EMC, very few persons in EMC R reported using the complete image name match strategy. To test this supposition, a future project could replicate the EMC R name recall training with follow up assessments at least one month after training. In theory, better self regulation might enhance training effects to the extent that it promotes persistent effort, when facing ch allenges. Partial support of this supposition comes from past research. Highly compliant EMC trainees who were more engaged with the training program, and completed more of the activities, gained more from the training (Bagwell & West, 2008). Further, in a working memory training paradigm, greater motivation to complete training, in terms of greater need for cognition or a growth mindset, was related to more training related gains (Jaeggi et al., 2014). In this particular study, relatively high intrinsic mo tivation across both training groups, or other gains between the training approaches. Individuals who are relatively disadvantaged in terms of motivation or baseline self regulatory factors might gain more from a beliefs demonstrate incremental benefit from this approach Such a moderation of training effects would be consistent with the role of memory self efficacy in goal setting and feedback effects for memory (West, Ebner & Hastings, 2013). This supposition could be tested by assessing motivational factors prior to training, pre selecting individuals with lower s cores on self regulation measures, or by manipulating motivation via varied participant recruitment or payment procedures (Jaeggi et al., 2014).
117 The most parsimonious explanation for the comparable training outcomes between the beliefs focused and strategy only training approaches is likely the limited training duration. Although many differences between the two training approaches were documented, the abbreviated nature of EMC R may not have allowed for sufficient presentation of elements that can enhance self regulation. Some elements to enhance self regulation are fundamental to any strategy training, such as repeated practice of skills and modeling of each strategy. Additional elements are inherent to training conducted in groups, such as having the whol e group practice strategies together or possible social support from other group members. Indeed, this may be why group training is consistently more effective in terms of memory gains than individual, self help training (Gross et al., 2012). To better und erstand the role of groups in beliefs focused training, future work might use self study individual training and active social groups as control conditions. In the present study the two groups differed in the extent that the instructor emphasized process improvement. There was also more general positive feedback during the beliefs focused classes. The beliefs focused group had a specific reading designed to help individuals adapt to age relate d memory changes. Training materials for the two groups also differed in the ordering of exercises, with easier first to encourage incremental mastery for the beliefs focused trainees. Even though this list indicates a large number of differences between t he two groups there are other possible elements to enhance self regulation that were limited here or were impractical to include within a single week. For instance, the longer EMC training program also included greater opportunities for in
118 class success us ing different strategies (including the learning of fellow classmate names) and small group discussions of homework questions to help trainees understand the homework readings. The most important of the EMC elements, dropped here due to time constraints, m ay have been specific goal setting for performance gains and positively setting and feedback may be critical for boosting training gains, given their impact on performance across adulthood (West et al., 2013). Comparison of training approaches with and without goal setting and provision of objective feedback on interim training gains is recommended, especially if investigators can identify a way to incorporate goal setting and feedback while still mainta ining a relatively short term training approach. Evidence for Near Transfer The most important contribution of EMC R to research on cognitive interventions for older persons may be the evidence for near transfer. That is, even though the EMC R strategies and practice exercises focused on learning and remembering name face pairs, trainees, but not control participants, demonstrated improved performance on the untrained, occupation name recall tasks (immediate and delayed). Of the four transfer tasks, correl ations between performance on these two and the targeted name recall task were larger than those between the occupation name delayed recognition task and the object location delayed recall task. Further, the two occupation name recall tasks were the most d ifficult of the transfer outcomes. Contrary to expectations, trainees in both the beliefs focused training condition and the strategy only training condition demonstrated pre post improvements in these transfer tasks. Regardless, evidence for near transfer in this present study is particularly important given the controversy surrounding generalization and transfer in cognitive interventions (Fisher, 2012) and the
119 decidedly limited evidence for transfer effects from past memory strategy training programs (Ze linski, 2009). Researchers agree that memory strategy training programs are effective in enhancing performance on trained tasks, immediately following training, but extensive debate surrounds the what, when, and how of transfer effects (Fisher, 2012). Cont overstated because the benefits are limited only to the trained tasks, and only immediately following training (McDaniel & Bugg, 2012). As such, demonstrating transfer to untraine d tasks that are similar to the targeted ability has nearly become a Evidence for transfer from memory strategy training is so rare that some scholars have proposed it should be circumvented by direct ly training the tasks that older persons want to enhance (McDaniel & Bugg, 2012 ), or by taking novel and alternative approaches to training, such as focusing on overall cognitive engagement (Stine Morrow & Basak, 2012; Stine Morrow et al., 2007; Rebok et a l., 2007). Some even suggest aborting the search for transfer from strategy training in favor of other types of cognitive interventions, such as those that offer extensive practice of core competencies (Zelinski, 2009 ). Indeed, transfer is more often obser ved with cognitive interventions that train core competencies, such as working memory, attention, or speed of processing, or that involve extensive practice of such skills, as in video game training (Jaeggi et al., 2008; Jaeggi et al., 2014; Kueider et al. 2012; Morrison & Chein, 2011; Toril et al., 2014). However, other scholars maintain that near transfer from memory strategy training is a reasonable goal, proposing that transfer might be promoted with
120 support (Fisher, 2012). Supports might include enhan ced monitori ng skills (Dunlosky et al., 2007 ; Dunlosky et al., 2003), encouragement to apply the learned strategies to new tasks (Cavallini et al., 2011), or changing the modality or location of training to more closely match the intended transfer settings e.g., offering training at home to promote An alternative solution may be to design memory strategy training programs using a theoretical model to foster transfer and this may be a primary reason fo r EMC issue in identifying transfer effects, across varied disciplines, is poor definition of what constitutes transfer and why. They proposed a taxonomy of classifying tr ansfer according to the content (what is transferred) and the context (the when and where of the learning and the transfer), and this taxonomy has been applied to a review of memory training for older adults (Zelinski, 2009). Regarding the context of trans fer, assessment outcomes (e.g., accuracy, speed) are identical (e.g., bo th speed or reaction time; both accuracy based), and (3) when the successful implementation of the trained procedures to the transfer task require minimal memory demand. EMC R met some of these criteria that are theorized to promote transfer. First, the tr principles, because the course covered how to enhance attention, knowledge about how memory works, and important age related memory changes. And, EMC R also pecific strategies, such as the image name match method for
121 learning names. Second, both the target and the transfer assessments, as well as the practice exercises, focused primarily on accuracy as the outcome, although speed was introduced as a secondary outcome for some of the more difficult practice exercises that were timed. Taken a step further, both the target task and the transfer tasks required associative memory skill, which relies on the ability to bind two pieces of information together. This ski ll often declines in late life (Naveh Benjamin, Guez, Kilb, & Reedy, 2004). Notably, near transfer in EMC R was suggested for the occupation name recall tasks (immediate and delayed), which were more similar (per intercorrelation coefficients) to the targe ted name recall task than the occupation name delayed recognition task or the object location delayed recall task. However, for EMC R successful application of the trained procedures to the transfer task did challenge memory. Participants would have neede d to recall the strategy that they learned while evaluating its effectiveness for the new task at hand. Thus, transfer may have occurred because considerable thought went into the design of the transfer tasks, as well as the development of appropriate mult ifactorial elements for the training program. The self regulatory gains associated with EMC R training may also deserve credit for the successful near transfer evidenced by EMC R trainees. Trainees in both the beliefs focused and strategy only training gro ups reported enhanced self evaluative beliefs and more effective strategy use following training. Likely these concurrent self regulatory benefits from training enabled trainees to adapt their approach to the associative memory testing in a way that led to improved perf ormance on the untrained tasks. Modest empirical evidence and theory suggest that enhanced self regulation might promote near transfer. For example, individuals trained to use self monitoring for
122 a specific memory task can effectively transfe r the use of this self regulatory strategy to other memory tasks, with instruction (Cavallini et al., 2010). Bandura (1997) posits that self regulation is key to task motivated behavior, task persistence, and, ultimately, task success. Reasonably, if an EM C R trainee felt more confident in her ability to complete memory activities, she might maintain effort longer on the challenging occupation name task, than if she did not feel confident in her capability to succeed on this task. Similarly, the flexible ad aptation and performance monitoring that likely underlies more effective strategy use might easily translate to different memory tasks, as long as the trainee has knowledge of strategies she might employ. Theoretically then, greater effectiveness of self regulatory system overall could contribute generally to task success, including performance on untrained memory tests. Limitations While the results of this research are quite promising, some possible limitations will need to be addressed in future r esearch. Sample selectivity may restrict the extent to which study results might be generalized. The sample was selective in terms of sex distribution, level of education and self reported health, and, potentially, motivation to complete the study. Additio nally, results of the study cannot address long term trajectories of training impact. The clear majority of the study sample was female participants. Well documented sex differences evidence that women outperform men on episodic recall, name recall, and fa ce recognition in mid life and old age (de Frias, Nilsson, & Herlitz, 2007; West, Crook, & Barron, 1993). Because the proportion of female participants were comparable across the three experimental conditions, these sex differences for memory were likely
123 n ot a confound for group differences. However, future research is needed to confirm whether the training benefits generalize to middle aged and older men. The participants in the present study were also relatively well educated and healthy, compared to the general population of middle aged and older Americans. Current debate in cognitive intervention literature questions who benefits the most from training. Some scholars propose a Matthew Effect, wherein those persons who are already relatively advantaged, in regards to cognitive capacity, education, or health, benefit most. Alternatively, greatest gains might be evidenced by relatively disadvantaged persons who have more to gain (Hertzog et al., 2009; Stine Morrow & Basak, 2011). Results of this research do suggest that relatively advantaged persons in terms of education and self reported health can benefit from brief strategy training for name recall. However, future research is needed to determine whether people with relatively low levels of education or b elow average health would benefit as much or more from this sort of training program. Participant recruitment materials advertised an everyday memory training program, and participant payments were the training materials. Thus, participants likely self se lected based on interest in improving their memory abilities. Considering this, the study sample may have had greater subjective complaints or more concerns about memory performance than the general population. Such selectivity would be concerning because subjective memory complaints are related to objective memory performance and may be a precursor to cognitive decline (Crumley et al., 2014; Reid & MacLullich, 2006). However, pretest memory self efficacy and general memory evaluation scores for this study were consistent with baseline values observed across six non training
124 memory studies from the same geographic area in Florida (cf. West, Bagwell, & Dark Freudeman, 2005; West et al., 2009; West, et al., 2003; West, Welch, & Thorn, 2001; West & Yassuda, 200 4). These data suggest that the sample was not disadvantaged in terms of memory complaints. Further because participants in the three different training conditions did not differ in baseline measures of working memory, episodic memory, or general memory e valuation, this selectivity if present would not have confounded comparisons made between these groups. The study recruitment and compensation procedures may have also attracted participants with high levels of intrinsic motivation to complete the trai ning program, which could have further supported training gains. The low attrition rate in this study, which was unrelated to experimental condition, further suggests that participants may have had high intrinsic motivation to complete the training. Across two working memory interventions, Jaeggi and colleagues (2014) evidenced that intrinsic motivation may promote transfer of training. This training program also provided novel evidence of near transfer, and the sample may have been intrinsically motivated to complete the training. However, whenever participants are recruited for training, per se, motivation could interact with training effects. If motivation is related to recruitment, then motivation should have also been present in the control group which did not show changes comparable to the trained group on most variables. Future work could directly assess or manipulate participant motivation to examine its effect on the practical impact of training. This research provides key evidence regarding the prac tical impact of training by demonstrating training related gains on trained and untrained memory tasks, as well as
125 improvements in self regulatory factors, immediately following training. However, because assessments were only administered in the weeks imm ediately before and immediately after training, data from this research cannot evaluate longer term trajectories of training related change. Cognitive researchers and aging persons alike are concerned with temporal transfer of training and question whether training related gains are sustained, and, if so, for how long (Hertzog et al., 2009), and evidence for long term training benefits is limited (Bottiroli et al., 2008; Rebok et al., 2014). Training related gains from the present study may not have persist ed beyond the week, or they might have been sustained or even magnified, and the long term trajectory of training effects may differ between the beliefs focused training approach and the strategy only training approach. These possibilities should be tested in future work with long term follow up assessments. Conclusion Collective results from cognitive intervention research demonstrate that middle aged and older persons may show test score improvements with small amounts of training (Gross et al., 2012; Her tzog et al., 2009), but researchers and trainees alike are concerned with the true practical impact of these programs. Researchers debate regulation how they evaluate their abilities and approach the task at hand as this could promote self sustaining performance gains (Hertzog & Dunlosky, 2012; Strickland Hughes & West 2017). Controversy also surrounds the scope of training benefits: are gains limited to performance on the targeted tasks? Individuals seeking train ing wonder whether they will be able to perceive a benefit from training and if the time or effort required is reasonable. EMC R provides insight towards resolution of these controversies with its tentative but optimistic
126 evidence that short term strategy training can enhance self regulatory factors and that benefits might generalize to similar, untrained tasks. When considering interventions for memory and aging, it might be tempting to focus on improved memory performance as an ultimate goal. Consistent w ith past programs, EMC R was effective in this goal, as performance on the trained memory task was improved for trainees compared to inactive control participants. However, performance is determined not solely by ability level but rather by the success of a comprehensive self regulatory system including personal, behavioral, and environmental factors. As such, the true success of EMC R was that the training enhanced specific self regulatory factors, namely memory self efficacy and effective strategy use. Th ese factors may be key to sustained memory success as they are related to current and future memory performance. Other training research has suggested that enhanced self regulation may be the special ingredient necessary for sizzling near transfer effects (Jaeggi et al., 2014; Payne et al., 2012; Strickland Hughes & West 2017). Certainly, results from this study support this supposition, as the training program benefited important self regulatory factors, above and beyond enhancing memory performance, and the program also provided novel evidence for near transfer. This research addresses contemporary controversy in the cognitive intervention literature by providing novel evidence that the practical impact of strategy training programs may generalize beyond the specific targeted task for middle aged and older adults. Evidence of n ear transfer to other cognitive tasks have been depressingly limited, with the majority of evidence produced by interventions focused on extended practice, such as core skill training or training via computer or video games (Strickland
127 Hughes & West, 2017) Thus, the modest evidence of near transfer in EMC R is highly remarkable and may be the most meaningful finding from this research. The direct value of strategy training for a middle aged or older trainees is immeasurably heightened if the benefits of th e program are not limited to the specifically trained task. Although speculative and future research is necessitated, near transfer, if linked to self regulatory gains, might extend to self sustaining and perpetuating performance gains. Replication of this finding is warranted, as well as more comprehensive examination of specific training characteristics and individual differences that might further promote transfer. Past research demonstrates that performance on targeted memory tests can improve with brie f training (Gross et al., 2012; Rebok & Balcerak, 1989), but EMC abbreviated training dosage is of great value, given its benefits to self regulation and near transfer. EMC R included approximately five hours of training, with only two of those hours i n person and instructor led. From a practical standpoint, this training could thus be easily incorporated into existing programming, such as those at Senior Centers or in lifelong learning programs, unlike the longer programs that are typical in the litera ture (providing 10 to 20 hours of group strategy training) Further, the low cost of a single session program in terms of both trainer time and participant time may incentivize initial participation and full completion of the training for middle aged a nd older adults. These benefits are important because memory training cannot be effective unless it can be offered and completed. Collectively, EMC R underscores the value of brief memory training for middle aged and older persons. Five hours of training o ver a single week can not only enhance performance on the trained memory task but also improve self regulation and untrained
128 memory performance. By focusing on memory strategies and self regulation in training, the benefits from the training program could translate from the laborat everyday life. A lifestyle aiding in many functional everyday activities and, ultimately, continuing their capacity to live independently.
129 APPENDIX A STRATEGY CHECKLIST S Techniques or Strategies for Recalling Names for Faces Some people are able to use special techniques to help them to remember. Here is a list of memory techniques. We would like to find out the methods that you used. You may have concentrated on the faces and did not do anything else. Or you may have tried many different strategies to try to learn the names. Either way is fine. We are interested only in finding out exactly what you did while you were studying. Please place a checkmark by all of the methods that you used while you were studying the names and faces. 1. I concentrated and paid attention to the face. 2. I concentrated and paid attention to the name. 3. I tried to pick out prominent features. 4. I tried to think of a meaningful association for the name. 5. I repeated names over and over to myself. 6. 7. I associated the face with the face of someone else I know. 8. I associated the name with the name of someone else I know. 9. be frank and honest 10. I m red Robin goes bob bob cheeks). 11. I created an image in my mind of the face. 12. I created an image in my mind of the face with a prominent facial feature exaggerated. 13. I created an image in my mind of the person with the name written across his or her face. 14. I created an image in my mind of the person next to an object that fits the name, 15. I used the image name match method. I created an image in my mind of the person, identified a concrete object related to the name, and I imagined that object next ture. 16. I covered the faces, looked away and tested myself on the names. 17. Other method(s): Please describe. ______________________________ ______ Now go back through and review the list of things you checked. Circle the 1 or 2 methods that you used the most often for remembering.
130 Techniques or Strategies for Recalling Locations of Objects Some people are able to use special techniques to help them to remember. Here is a list of memory techniques. We would like to find out the methods that you used. You may have put the objects where you would in your own house and did not do anything else. Or you may have tried many different strategies to try to remember the object locations. Either way is fine. We are interes ted only in finding out exactly what you did while you were studying. Please place a checkmark by all of the methods that you used while you were studying the object locations. 1. I concentrated and paid attention to the rooms. 2. I concentrated and paid attention to the objects. 3. I concentrated on how the objects looked in the rooms. 4. I repeated the locations of objects over and over. 5. I paired the first letter of the object with the first letter of the room. 6. I put the objects in logical rooms, where most people would expect to find those objects. 7. I put two objects together in a room because the two objects logically go together (e.g., a lamp and table). 8. I put individual objects where I put those same objects in my own house. 9. I put two o bjects together because I have those objects placed together in the same room in my own house. 10. I made up a sentence about the objects, one at a time, which matched each object to its room. 11. I made up a story using many of the objects in their rooms. 12. I created simple mental images of the objects in the rooms. 13. 14. I made active mental pictures, like a movie, that showed the objects in the rooms. 15. I closed my eyes, thought of the different rooms shown and checked whether I could list the object(s) I placed in each room. 16. Other method(s): Please describe. ____________________________________ _______________________________________________________ ____________ ___________________________________________________________________ Now go back through and review the list of things you checked. Circle the 1 or 2 methods that you used the most often for remembering
131 Techniques or Strategies for Recallin g Names for Occupations people are able to use special techniques to help them to remember. Here is a list of memory techniques. We would like to find out the methods that you used. You may have concentrated on the occupations and did not do anything else Or you may have tried many different strategies to try to learn the names. Either way is fine. We are interested only in finding out exactly what you did while you were studying. Please place a checkmark by all of the methods that you used while you were studying the occupations and names. 1. I concentrated and paid attention to the occupation. 2. I concentrated and paid attention to the name. 3. I repeated names over and over to myself. 4. I repeated occupations over and over to myself. 5. I repeated names and occupations together. 6. I tried to think of a meaningful association for the name. 7. 8. I tried to think about the nature o f the occupation, e.g., barbers cut hair with scissors; judges are wise and wear robes. 9. 10. I associated the occupation with someone I know who has that same occupation. 11. I associated the name with the name of someone else I know. 12. don (the lawyer) wears a gor geo 13. I created an image in my mind related to the nature of the occupation. 14. I created an image in my mind of a person with the occupation, and wrote the 15. I created an image in my mind of a pers on with the occupation next to a concrete 16. I covered the occupations, looked away and tested myself on the names. 17. Oth er method(s). Please describe: ____________________________________ ___________________________________________________________________ ___________________________________________________________________ Now go back through and review the list of things yo u checked. Circle the 1 or 2 methods that you used the most often for rememb ering.
132 APPENDIX B OJBECT LOCATION VISUAL ASSOCIATION MATERIALS Table B 1. List of objects used in the object location visual association task Set A Objects Set B Objects bottle legal pad ashtray knife box light bulb banner magnifying glass cigars luggage baseball paper bag clock milk blender pen comb paint bucket calculator phone dress passport calendar pillow dumbbell perfume camera purse envelope radio cassette tape screwdriver fire extinguisher ring crutches shoes flag stapler football teapot hole puncher tennis racket hand axe trashcan ladder thumb tack horn wrench
133 Figure B 1. Example objects used in the object location visual association task
134 Figure B 2 Example matrix array used in object location visual association task
135 APPENDIX C NAME OCCUPA TION VERBAL ASSOCIATION TASK MATERIALS Table C 1. List of occupation name pairs by set Set A Pairs Set B Pairs Occupation Name Occupation Name ARCHITECT ALLEN ACCOUNTANT POTTER CHEMIST BENNETT ACTOR JONES COMIC OLIVER ATHLETE STEVENS CONDUCTOR WEAVER BANKER CARPENTER DENTIST WAGNER BUILDER BANKS DESIGNER JEFFERSON BUTCHER HARRISON DETECTIVE ROGERS CLERK FOX DIPLOMAT SCOTT COACH WATERS DOCTOR HUNTER GARDENER MITCHELL DRIVER HOWE GROCER WILSON ENGINEER GRAHAM INVENTOR MORTON FIREMAN BUTLER LOGGER MONROE GUARD BISHOP MANAGER KING LAWYER GRAVES MERCHANT DIXON LIBRARIAN SHEPHERD PILOT CHURCH MECHANIC GILMORE PLUMBER WELLS MINER SHOEMAKER POLICEMAN DAVIS NURSE COOK POLITICIAN BREWER PAINTER WILLIAMS POPE FLOWERS PASTOR STANLEY POSTMAN GILBERT PRINCE MORGAN PRIEST BARBER PRINTER WALKER SALESMAN SPENCER RABBI BERRY SCIENTIST HARDING REPORTER BAKER SERGEANT HAMILTON RUNNER STONE SINGER FISHER SOLDIER FLYNN SURGEON PETERS TAILOR DOUGLAS TUTOR BROOKS TEACHER BEARD WAITER HOBBS USHER MEADOWS WELDER FARMER VET CURRY WRITER KNIGHT
136 Figure C 1. Occupation name verbal association task instructions
137 APPENDIX D HEALTH AND MEMORY ENGAGEMENT SURVEY Pretest Health and Memory Engagement Survey Please give us the following information about yourself. If needed, you can use more space on the back of this page. 1. Your date of birth: ___________________________ 2. Have you been hospitalized in the last year? If YES, please describe the reason(s) for each hospitalization. Circle one: YES or NO 3. Is your physician currently treating you for any specifi c illnesses or health problems? If YES, please list these conditions. Circle one: YES or NO 4. Please list any prescription medications you take regularly. If you DO NOT take any prescription medications regularly, please circle below. If you DO take something regularly, please give the names of the medications and describe your reasons for taking each one. If you do not know the name of a medication, just list the reason yo u take it. NONE 5. Other than vitamins, please list any over the counter or non prescription medications such as medicinal herbs, aspirin, cold medicines, etc that you have taken within the past 48 hours If you have not taken any non prescription NONE NONE
138 Please give us the follow ing information about yourself. If needed, you can use more space on the back of this page. Some memory training programs are offered with a personal instructor, others are offered online or with CDs/books. 6. Have you completed memory training with a personal instructor in the last year? If YES, please describe your training, including how long it took and the type of activities you did. Circle one: YES or NO 7. Have you completed memory training with guided instruction via computer, videos, or written materials, such as a workbook, in the past year? If YES, please describe your training, including how long it took and the typ e of activities you did. Circle one: YES or NO For 8 11, please indicate how often you engaged in each activity within the last year Do not worry if you cannot give an exact figure. Circle the letter that most nearly describes the frequ ency with which you have done the activity, using the scale below. a. Never f 2 or 3 times a month b. Less than once a year g. About once a week c. About once a year h. 2 or 3 times a week d. 2 or 3 times a year i. Daily e. About once a month 8. Exercise my memory with computer games or activities a b c d e f g h i 9. Exercise my memory with games or activities not on the computer a b c d e f g h i 10. Learn and practice new strategies for my everyday memory activities a b c d e f g h i 11. Practice well known strategies for my everyday memory activities a b c d e f g h i
139 Posttest Health and Memory Engagement Survey Please give us the following information about yourself. If needed, you can use more space on the back of this page. 1. Your date of birth: ___________________________ 2. Have you been hospitalized since your last interview two weeks ago ? If YES, please describe the reason(s) for each hospitalization. Circle one: YES or NO 3. Has your physician diagnosed you with a new illness since your last interview ? If YES, please list these conditions. Circle one: YES or NO 4. Please list any NEW prescription medications you have started taking since your last interview If you are not taking anything new please circle below. If you do not know the name of a medication, just list the reason you take it. NONE 5. Other than vitamins, please list any over the counter or non prescription medication s such as medicinal herbs, aspirin, cold medicines, etc that you have taken within the past 48 hours If you have not taken any non prescription NONE NONE
140 Please give us the follow ing information about yourself. If needed, you can use more space on the back of this page. Some memory training programs are offered with a personal instructor, others are offered online or with CDs/books. 6. Other than the Everyday Memory Training program h ave you completed memory training with a personal instructor since your last interview two weeks ago ? If YES, please describe your training, including how long it took and the type of activities you did. Circle one: YES or NO 7. Other than the Everyday Memory Training program h ave you completed memory training with guided instruction via computer, videos, or written materials, such as a workbook, since your last interview ? If YES, please describe your training, including how long it took and the type of activities you did. Circle one: YES or NO For 8 11, please indicate how often you engaged in each activity since your last interview Do not worry if you cannot give an exact figure. Circ le the letter that most nearly describes the frequency with which you have done the activity. Please consider activities you may have done for the Everyday Memory Training program as well as other activities. 8. Exercise my memory with computer games or activities never daily a b c d e f g h i 9. Exercise my memory with games or activities not on the computer never daily a b c d e f g h i 10. Learn and practice new strategies for my everyday memory activities never daily a b c d e f g h i 11. Practice well known strategies for my everyday memory activities never daily a b c d e f g h i
141 APPENDIX E HOMEWORK ASSIGNMENT SHEETS Homework Ass ignment Sheet for the Strategy Plus Beliefs Condition Homework Please complete all homework before your next interview session or within one week, whichever comes first. Record the date and time spent on each activity on your Activity Log The purpose of the homework: 1) Show you a wide variety of memory techniques that you could use 2) Help you to learn the recommended memory strategies presented in class Assignments Read Age and Memory. Read How to Remember Names Complete a minimum of 40 minutes of practice with the exercises in the workbook. Find a picture in your home and use Active Observation to remember the details in the picture. Answer the homework questions below. Please write your answers on the back of this page (use back of Activity Log if you need more space). 1. De scribe at least two age related memory changes. 2. What should you do if you need more cues to remember? 3. Identify the name learning technique that seems like the best strategy for you. Use this method to remember at least one name this week. Report on how you used the strategy. 4. What kinds of activities do you do at home that are mentally challenging, that require you to think? Write at least two examples. Do you have any questions about the homework assignments? If so, please contact us by emailing firstname.lastname@example.org or calling (352) 273 3829.
142 Homework Assignment Sheet for the Strategy Only Condition Homework Please complete all homework before your next interview session or within one week, whichever comes first. Record the date and time spent on each activity on your Activity Log The purpose of the homework: 1) Show you a wide variety of memory techniques that you could use 2) Help you to raise your scores by using the memory strategies from class Assignments Read How Memory Works Read How to Remember Names Complete a minimum of 40 min utes of practice with the exercises in the workbook. Find a picture in your home and use Active Observation to remember the details in the picture. Answer the homework questions (below). Please write your answers on the back of this page (use back of Acti vity Log if you need more space). 1. How are working memory and long term memory different? 2. How do memory cues work? 3. Use the sentence strategy from the reading on names to remember at least one name this week. Report on how you used the strategy. 4. What kinds of activities do you do at home that are mentally challenging, that require you to think? Write at least two examples. Do you have any questions about the homework assignments? If so, please contact us by emailing email@example.com or calling (3 52) 273 3829.
143 LIST OF REFERENCES Agrigoroaei, S., & Lachman, M. E. (2010). Personal control and aging: How beliefs and expectations matter. In J.C. Cavanaugh, C. K. Cavanaugh, J. M. Berry, and R. L. West, Eds., Aging in America, Volume 1: Psychological Aspects of Aging (pp. 177 201). New York: Praeger Press. Agrigoroaei, S., Neupert, S. D., & Lachman, M. E. (2013). Maintaining a sense of control in the context of cognitive challenge: Greater stability in control beliefs benefits working mem ory. The Journal of Gerontopsychology and Geriatric Psychiatry, 26 (1), 49 59. doi:10.1024/1662 9647/a000078 Bagwell, D. K., & West, R. L. (2008). Assessing compliance: Active versus inactive trainees in a memory intervention. Clinical Interventions in Aging, 3 (2). 371 382. doi:10.2147/CIA.S1413 Bailey, H., Dunlosky, J., & Hertzog, C. (2010). Metacognitive training at home: Does it Gerontology, 56 (4), 414 420. doi:10.1159/000266030 Ball, K., Berch, D. B., Helmers, K. F., J obe, J. B., Leveck, M. D., Marsiske, M., Morris, J. N., Rebok, G. W., Smith, D. M., Tennstedt, S. L., Unverzagt, F. W., & Willis, S. L. (2002). Effects of cognitive training interventions with older adults: A randomized controlled trial. Journal of the Ame rican Medical Association, 288 (18), 2271 2281. doi:10.1001/jama.288.18.2271 Ball, K., Edwards, J. D., & Ross, L. A. (2007). The impact of speed of processing training on cognitive and everyday functions. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 62 (Special Issue 1), 19 31. doi:10.1093/geronb/62.special_issue_1.19 Bandura, A. (1997). Self efficacy: The exercise of control New York: W. H. Freeman. Barnett, S. M., & Ceci, S. J. (2002). When and where do we apply what w e learn? A taxonomy for far transfer. Psychological Bulletin, 128 (4), 612 637. doi:10.1037/0033 2909.128.4.612 Beaudoin, M. & Desrichard, O. (2011). Are memory self efficacy and memory performance related? A meta analysis. Psychological Bulletin, 137 (2), 2 11 241. doi:10.1037/a0022106 Berry, J., Hastings, E., West, R., Lee, C., & Cavanaugh, J. C. (2010). Memory aging: Deficits, beliefs, and interventions. In J. C. Cavanaugh, C. K. Cavanaugh, J. Berry, & R. West (Eds.), Aging in America, Vol. 1. Psychological Aspects of Aging (pp. 255 299). Santa Barbara, CA: Praeger Press.
144 Blanchard Fields, F., Horhota, M., & Mienaltowski, A. (2008). Social context and cognition. In S. Hofer & D. Alwin (Eds.), Handbook on Cognitive Aging: Interdisciplinary Perspectives (pp. 614 628). Thousand Oaks, CA: Sage Publications. doi:10.4135/9781412976589 Boot, W. R., Simons, D. J., Stothard, C., & Stutts, C. (2013). The pervasive problem with placebos in psychology: Why active control groups are not sufficient to rule out placebo eff ects. Perspectives on Psychological Science, 8 (4), 445 454. doi:10.1177/1745691613491271 Bottiroli, S., Cavallini, E., & Vecchi, T. (2008). Long term effects of memory training in the elderly: A longitudinal study. Archives of Gerontology and Geriatrics, 4 7 (2), 277 289. doi:10.1016/j.archger.2007/08.101 Brandt, J., Spencer, M., & Folstein, M. (1988). The Telephone Interview for Cognitive Status. Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 1 111 117. Brown, K. W., & Ryan, R. M. (2003). The b enefits of being present: Mindfulness and its role in psychological well being. Journal of Personality and Social Psychology, 84 (4), 822 848. doi:10.1037/0022 35220.127.116.112 Carlson, M. C., Erickson, K. I., Kramer, A. F., Voss, M. W., Bolea, N., Mielke, M., McGill, S., Rebok, G. W., Seeman, T., & Fried, L. P. (2009). Evidence for neurocognitive plasticity in at risk older adults: The Experience Corps program. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences, 64 (12), 1275 1282. doi:10.1093/Gerona/glp117 Cavallini, E., Dunlosky, J., Bottiroli, S., Hertzog, C., & Vecchi, T. (2010 ). Promoting transfer in memory training for older adults. Aging, Clinical, and Experimental Research, 22 (4), 314 323. doi:10.3275/6704. Charness, N. (Ed.) (2007). Cognitive interventions and aging [Special issue]. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 62 (Special Issue 1). Chasteen, A. L., Kang, S. K., & Remedios, J. D. (2011). Aging and stereotype threat. In M. I nzlicht & T. Schmader (Eds.), Stereotype threat: Theory, process, and application (1st edition; pp. 202 216). New York: Oxford University Press. Cohen Mansfield, J., Cohen, R., Buettner, L., Eyal, N., Jakobovits, H., Rebok, G., Rotenberg Shpigelman, S., & Sternberg, S. (2014). Interventions for older persons reporting memory difficulties: A randomized controlled pilot study. International Journal of Geriatric Psychiatry. Advance online publication. doi:10.1002/gps.4164 Craigie, M., & Hanley, J. R. (1997). P utting faces to names. British Journal of Psychology, 88 (1), 157 171. doi:10.1111/j.2044 8295.1997.tb02626.x
145 Crumley, J. J., Stetler, C. A., & Horhota, M. (2014). Examining the relationship between subjective and objective memory performance in older adult s: A meta analysis. Psychology and aging, 29 (2), 250 263. doi:10.1037/a0035908 Dark Freudeman, A., West, R. L., & Viverito, K. (2006). Future selves and aging: Older Educational Gerontology, 32 85 109. doi:10.1080/0360127050038 8125 de Frias, C. M., Nilsson, L G., & Herliz, A. (2006). Sex differences in cognition are stable over a 10 year period in adulthood and old age. Aging, Neuropsychology, and Cognition, 13 (3 4(, 574 587. doi:10.1080/13825580600678418 Dixon, R. A., Hultsch D. F., & Hertzog, C. (1988). The Metamemory in Adulthood (MIA) questionnaire. Psychopharmacology Bulletin, 24 671 688. doi:10.1037/t02356 000 Dunlosky, J., Cavallini, E., Roth, H., McGuire, C. L., Vecchi T., & Hertzog, C. (2007). Do self monitoring interventions improve older adults learning? The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 62 (Special Issue 1). 70 76. Dunlosky, J., Kubat Silman, A., & Hertzog, C. (2003). Training metacognitive skills paced associative learning. Psychology and Aging, 18 (4), 340 345. doi:10.1037/0882 7918.104.22.1680 Faul, F., Erdfelder, E., Lang, A G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39 (2), 175 191. doi:10.3758/bf03193146 Federal Interagency Forum on Aging Related Statistics. (2012). Older Americans 2012: Key Indicators of Well Being Feder al Interagency Forum on Aging Related Statistics. Washington, DC: U.S. Government Printing Office. Fisher, R. (Ed.) (2012). Target article and commentaries. Journal of Applied Research in Memory and Cognition, 1 (1). doi:10.1016/j.jarmac.2012.01.001 Floyd, M., & Scogin, F. (1997). Effects of memory training on the subjective memory functioning and mental health of older adults: A meta analysis. Psychology and Aging, 12 (1), 150 161. doi:10.1037/0882 7922.214.171.124 Fried, L. P., Carlson, M., Freedman, M., Frick K. D., Glass, T. A., Hill, J., McGill, S., Rebok, G. W., Seeman, T., Tielsch, J., Wasik, B. A., & Zeger, S. (2004). A social model for health promotion for an aging population: Initial evidence on the Experience Corps model. Journal of Urban Health, 81 (1 ), 64 78. doi:10.1093/jurban/jth094
146 Gross, A. L., Parisi, J. M., Spira, A. P., Kueider, A. M., Ko, J. Y., Saczynski, J. S., Samus, Q. M., & Rebok, G. W. (2012). Memory training interventions for older adults: A meta analysis. Aging and Mental Health, 16 (6 ), 722 734. doi:10.1080/13607863.2012.667783 Gross, A. L., & Rebok, G. W. (2011). Memory training and strategy use in older adults: results from the ACTIVE study. Psychology and Aging, 26 (3), 503 517. doi:10.1037/a0022687 Hastings, E. C., & West, R. L. (20 09). The relative success of a self help and a group based memory training program for older adults. Psychology and Aging, 24 (3), 586 594. doi:10.1037/a0016951 Hastings, E. C., & West, R. L. (2011). Goal orientation and self efficacy in relation to memory in adulthood. Aging, Neuropsychology, and Cognition, 18 (4), 471 493. doi:10.1080/13825585.2011.575926 Hauser, R. M., & Willis, R. J. (2005). Survey design and methodology in the Health and Retirement Study and the Wisconsin Longitudinal Study. Population a nd Development Review, 30 (Supplement: Aging, Health, and Public Policy), 209 235. doi: 10.7826/isr um.06.585031.001.05.0012.2005 Hayes, S. M., & Cabeza, R. (2008). Imaging aging. In S. M. Hofer & D. F. Alwin (Eds.), The Handbook on Cognitive Aging: Interdis ciplinary Perspectives (pp. 308 326). Thousand Oaks, CA: Sage Publications. Hering, A., Rendell, P. G., Rose, N. S., Schnitzspahn, K. M., & Kliegel, M. (2014). Prospective memory training in older adults and its relevance for successful aging. Psychologica l Research, 78 (6), 892 904. doi:10.1007/s00426 014 0566 4 Hertzog, C. & Dunlosky, J. (2011). Metacognition in later adulthood: Spared monitoring regulation. Current Directions in Psychological Science, 20 (3), 167 173. doi:10. 1177/096372141140926 Hertzog, C. & Dunlosky, J. (2012). Metacognitive approaches can promote transfer of training: Comment on McDaniel and Bugg. Journal of Applied Research in Memory and Cognition, 1 (1). d oi:10.1016/j.jarmac.2012.01.003 Hertzog, C., & Hul tsch, D. F. (2000). Metacognition in adulthood and old age. In F. I. M. Craik & T. A. Salthouse (Eds.), The Handbook of Aging and Cognition (2 nd ed., pp. 417 466). Mahwah, NJ: Erlbaum. Hertzog, C., & Jopp D. S. (2010). Resilience in the face of cognitive aging: Experience, adaptation, and compensation. In P. S. Fry, & C. Keyes (Eds.), New Frontiers in Resilience Aging: Life Strengths and Wellness in Late Life (pp. 130 161). NY: Cambridge University Press.
147 Hertzog, C., Kramer, A. F., Wilson, R. S., & Lindenberger, U. (2009). Enrichment effects on adult cognitive development: Can the functional capacity of older adults be preserved and enhanced? Psychological Science in the Public Interest, 9 (1), 1 65. doi:1 0.1111/j.1539 6053.2009.01034.x Hummert, M. L. (2011). Age stereotypes and aging. In K. W. Schaie & S. L. Willis, Handbook of the psychology of aging (7th ed., pp. 249 262). London: Elsevier. IBM Corp. (2016). IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp. Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008). Improving fluid intelligence with training on working memory. PNAS, 105 (19), 6829 6833. doi:10.1073/pnas.0801268105 Jaeggi, S. M., Buschkuehl, M., Shah, P., & Jonide s, J. (2014). The role of individual differences in cognition training and transfer. Memory and Cognition, 42 (3). 464 480. doi:10.3758/s13421 013 0364 z James, L. E. (2004). Meeting Mr. Farmer versus meeting a farmer: Specific effects of aging on learning proper names. Psychology and Aging, 19 (3), 515 522. doi:10.1037/0882 7974.19.3515 Kastenbaum, R., Derbin, V., Sabatini, P., & Artt, S. (1972). The ages of me: Toward personal and interpersonal definitions of functional aging. Aging and Human Development, 3 (2) 197 211. doi:10.2190/TUJR WTXK 866Q 8QU7 Kliegel, M. & Brki, C. (2012). Memory training interventions require a tailor made approach: Commentary on McDaniel and Bugg. Journal of Applied Research in Memory and Cognition, 1 (1). doi:10.1016/j.jarmac.201 2.01.002 Kueider, A. M., Parisi, J. M., Gross, A. L., & Rebok, G. W. (2012). Computerized cognitive training with older adults: A systematic review. PLoS ONE, 7 (7), e40588, doi:10.1371/journal.pone.0040588 Lachman, M. E., Agrigoroaei, S., Tun P. A., & Weaver, S. L. (2014). Monitoring cognitive functioning: Psychometric properties of the Brief Test of Adult cognition by Telephone. Assessment, 2 (4), 404 417. doi:10.1177/1073191113508807 Lachman, M. E., Neupert, S. D., & Agrigoroaei, S. (2011). The relevance of control beliefs for health and aging. In K. W. Schaie & S. L. Willis (Eds.), Handbook of the Psychology of Aging (7 th ed., pp. 175 190). New York, NY: Elsevier. Lachman, M. E., & Weaver, S. L. (1998). The sense of control as a moderator of social class differences in health and well being. Journal of Personality and Social Psychology, 74 (3), 763 773. doi:10.1037/0022 35126.96.36.1993
148 Lachman, M. E., Weaver, S. L., Bandura, M., Elliott, E., & Lewkowitz, C. J. (1992). Improving memory and contro l beliefs through cognitive restructuring and self generated strategies. The Journals of Gerontology, Series B: Psychological and Social Sciences, 27 (5), 293 299. doi:10.1093/geronj/47.5.p293 Lezak, M. D. (1995). Neuropsychological assessment, 3rd edition. New York: Oxford University Press. Mather, M. (2010). Aging and cognition. Wiley Interdisciplinary Reviews: Cognitive Science, 1 (3), 346 362. McDaniel, M. A., & Bugg, J. (2012). Memory training interventions: What has been forgotten? Journal of Applied R esearch in Memory and Cognition, 1 (1). doi:10.1016/j.jarmac.2012.11.002 McDougall, G. J., Becker, H., PItuch, K., Acee, T. W., Vaughan, P. W., & Delville, C. L. (2010). The Senior WISE study: Improving everyday memory in older adults. Archives of Psychiatric Nursing, 24 (5), 291 306. doi:10.1016/j.apnu.2009.11.001 Montegjo, P., Montenegro, M., Fernndez, M. A., & Maest, F. (2012). Memory complaints in the elderly: Quality of life and daily living activities. A population based study. Archives of Ge rontology and Geriatrics, 54 298 304. doi:10.1016/j.archger.2011.05.021 Morrison, A., & Chein, J. M. (2011). Does working memory training work? The promise and challenges of enhancing cognition by training working memory. Psychonomic Bulletin and Review, 18 (1), 46 60. doi:10.3758/s13423 010 0034 0 Naveh Benjamin, M., Guez, J., Kilb, A., & Reedy, S. (2004). The associative memory name associations. Psychology and Aging, 19 (3), 541 546. doi:10.1037/0882 797 188.8.131.521 Nelson, T. O. & Narens, L. (1990). Metamemory: A theoretical framework and new f indings. Psychology of Learning and Motivation 125 173. doi:10.1016/s0079 7421(08)60053 5 Noice, H., & Noice, T. (2013). Extending the reach of an evidence based th eatrical intervention. Experimental Aging Research, 39 (4), 398 4180 doi:10.1080/0361073X.2013.808116 Nyberg, L., Lvdn, M., Riklund, K., Lindenberger, U., & Bckman, L. (2012). Memory aging and brain maintenance. Trends in Cognitive Sciences, 16 (5), 292 3 05. doi:10.1016/j.tics.2012.04.005 Open Science Collaboration (2015). Estimating the reproducibility of psychological science. Science, 349 (6251), aac4716 1 aac4716 9. doi: 10.1126/science.aac4716
149 Ownby, R. L., Crocco, E., Acevedo, A., Vineeth, J., & Loewen stein, D. (2006). Depression and risk for Alzheimer disease: Systematic review, meta analysis, and metaregression analysis. Archives of General Psychiatry, 63 (5), 530 538. doi:10.1001/archpsyc.63.5.530 Park, D. C., & Bischof, G. N. (2013). The aging mind: Neuroplasticity in response to cognitive training. Dialogues in Clinical Neuroscience, 15 (1), 109 119. Payne, B. R., Jackson, J. J., Hill, P. L., Gao, X., Roberts, B. W., & Stine Morrow, E. A. L. (2012). Memory self efficacy predicts responsiveness to indu ctive reasoning training in older adults. The Journals of Gerontology, Series B: Psychological and Social Sciences, 67 27 35. doi:10.1093/geronb/gbr073 Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19 (1), 2 21. doi:10.2307/2136319 Rapp, S., Brenes, G., & Marsh, A. P. (2002). Memory enhancement training for older adults with mild cognitive impairment: A preliminary study. Aging and Mental Health, 6 (1), 5 11. doi:10.1080/13607860120101077 Rebok, G. W., & Balcerak, L. J. (1989). Memory self efficacy and performance differences in young and old adults: Effect of mnemonic training, Developmental Psychology, 25 (5), 714 721. doi: 10.1037/0012 16184.108.40.2064 Rebok, G. W., Ball, K., Guey, L. T., Jones, R. N., Ki m, H Y., King, J. W., Marsiske, M., Morris, J. N., Tennstedt, S. L., Unverzagt F. W., & Willis, S. L. (2014). Ten year effects of the Advanced Cognitive Training for Independent and Vital elderly Cognitive Training trial on cognition and everyday functioning in older adults. Journal of the American Geriatrics Society, 62 (1), 16 24. doi:10.1111/jgs.12607 Rebok, G. W., Carlson, M. C., & Langbaum, J. B. S. (2007). Training and maintaining memory abilities in healthy older adults: Traditional and novel approaches. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences,62 (Special Issue), 53 61. doi:10.1093/geronb/62.special_issue_1.53 Rebok, G. W., Langbaum, J. B. S., Jones, R. N., Gross, A. L., Parisi, J. M., Spira, A. P., Kueider, A., Petras, H., & Brandt, J. (2013). Memory training in the ACTIVE study: How mu ch is needed and how benefits? Journal of Aging and Health 25 (8 Suppl), 21S 42S. doi:10.1177/0898264312461937 Reid, L. M., & MacLullich, A. M. J. (2006). Subjective memory complaints and cognitive impairment in older people. Dementia and Geriatric Cogniti ve Disorders, 22 (5 6), 471 485. doi: 10.1159/000096295 Salthouse, T. A. (2006). Mental exercise and mental aging: Evaluating the validity of the Perspectives on Psychological Science, 1 (1), 68 87. doi:10.1111/j.1745 6916.2006 .00005.x
150 Scogin, F., & Bienias, J. L. (1988). A three year follow up of older adult participants in a memory skills training program. Psychology and Aging, 3 (4), 334 337. doi:10.1037/0882 79220.127.116.114 Sisco, S., Marsiske, M., Gross, A. L., & Rebok, G. W. ( 2013). The influence of cognitive Journal of Aging and Health, 25 (8S), 230S 248S. doi:10.1177/0898264313501386 Smith, S. L., & West, R. L. (2006). The application of self efficacy principles to audiologic rehabilitation: A tutorial. American Journal of Audiology, 15, 46 56. doi:10.1044/1059 0889(2006/006) Stigsdotter Neely, A., & Bckman, L. (1993). Long term maintenan ce of gains from memory training in older adults: Two 3 year follow up studies. Journal o f Gerontology: Psychological Sciences, 48 (5), 232 237. doi:10.1093/geronj/45.8.P233 Stine Morrow, E. A. L., & Basak, C. (2011). Cognitive interventions. In K. W. Schaie & S. L. Willis (Eds.), Handbook of the Psychology of Aging (7 th ed., pp. 153 170). New York: Elsevier. Stine Morrow, E. A. L., Parisi, J. M., Morrow, D. G., Greene, J., & Park, D. C. (2007). An engagement model of cognitive optimization through adulthood. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences,62 (Sp ecial Issue), 62 69. doi:10.1093/geronb/62.special_issue_1.62 Stine Morrow, E. A. L., Payne, B. R., Hill, P., Jackson, J., Roberts, B., Kramer, A., Morrow, D. G., Payne, L., Janke, M., Noh, S. R., & Parisi, J. M. (2014). Training versus engagement as paths to cognitive optimization with aging. Psychology and Aging, 29 (4), 891 906. doi:10.1037/a0038244 Strickland Hu ghes, C. M., & West, R. L. (2017 ). Memory: Training methods and benefits. In N. A. Pachana (Ed.) Encyclopedia of Geropsychology New York, NY: Sp ringer. Strickland Hughes, C. M., West, R. L., Smith, K. A., & Ebner, N. C. (2016). False feedback and beliefs influence name recall in younger and older adults. Memory doi:10.1080/09658211.2016.1260746 Toril, P., Reales, J. M., & Ballesteros, S. (2014). Video game training enhances cognition of older adults: A meta analytic study. Psychology and Aging, 29 (3), 706 716. doi:10.1037/a0037507 Valentijn, S. .A. M., van Hooren, S. A. H., Bosma, H., Touw, D. M., Jolles, J., van Boxtel, M. P. J., & Ponds, R. W. H M. (2005). The effect of two types of memory training on subjective and objective memory performance in healthy individuals aged 55 years and older: A randomized controlled trial. Patient Education and Counseling, 57 (1), 106 114. doi:10.1015/j.pec.2005.0 5.002
151 Valentijn, S. A. M., Hill, R. D., Van Hooren, S. A. H., Bosma, H., Van Boxtel, M. P. J., Jolles, J., & Ponds, R. W. H. M. (2006). Memory self efficacy predicts memory performance: Results from a 6 year follow up study. Psychology and Aging, 21 165 1 72. doi:10.1037/0882 7918.104.22.168 Verhaeghen, P., & Marcoen, A. (1996). On the mechanisms of plasticity in young and older adults after instruction in the method of loci: Evidence for an amplification model. Psychology and Aging, 11 (1), 164 178. doi:10.10 37/0882 7922.214.171.124 Verhaeghen, P., Marcoen, A., & Goossens, L. (1992). Improving memory performance in the aged through mnemonic training: A meta analytic study. Psychology and Aging, 7 (2), 242 251. doi:10.1037/0882 79126.96.36.199 Ware, J. E., Kos inski, M., & Keller, S. D. (1996 ). A 12 item short form health survey: construction of scales and preliminary tests of reliability and validity. Medical Care, 34 (3), 220 233. doi:10.1097/00005650 199603000 00003 Ware, J. E., Kosinski, M., & Keller, S. D. (1995). SF 12: How to score the SF 12 physical and mental health summary scales (2 nd ed.). Boston, MA: The Health Institute, New England Medical Center. Wechsler, D. (1997). Wechsler Adult Intelligence Scale Third Edition (WAIS). New York, NY: Psychological Corpor ation. Welch, D. C. (1998). The effects of object location training on spatial self efficacy, location recall, and strategy usage (Unpublished doctoral dissertation). University of Florida, Gainesville, FL. West, R. L. (1985). Memory fitness over 40 Gaine sville, FL: Triad Publishing Company. West, R. L. (1989). Planning practical memory training for the aged In L. W. Poon, D. C. Rubin, & B. D. Wilson (Eds.), Everyday Cognition in Adulthood and Late Life (pp. 573 597). Cambridge, UK: Cambridge University P ress. doi:10.1017/cbo9780511759390.031 West, R. L. (2012). A commentary on historical and innovative approaches to memory intervention for older adults. Journal of Applied Research in Memory and Cognition, 1 (1). doi:10.1016/j.jarmac.2012.12.003 West, R. L. Bagwell, D., & Dark Freudeman, A. (2005). Memory and goal setting: The response of older and younger adults to positive and objective feedback. Psychology and Aging, 20 (2), 195 201. doi:10.1037/0882 79188.8.131.52 West, R. L., Bagwell, D., & Dark Freudema n, A. (2008). Self efficacy and memory aging: The impact of a memory intervention based on self efficacy. Aging, Neuropsychology, and Cognition, 15 302 329. doi:10.1080/13825580701440510
152 West, R. L., Crook, T. H., & Barron, K. L. (1992). Everyday memory p erformance across the life span: Effects of age and noncognitive individual differences. Psychology and Aging, 7 (1), 72 82. West, R. L., Dark Freudeman, A., & Bagwell, D. (2009). Goals feedback conditions and memory: Mechanisms for memory gains in older an d younger adults. Memory, 17 (2), 233 244. doi:10.1080/09658210802236407 West, R. L., Ebner, N. C., & Hastings, E. C. (2013). Linking goals and aging: Experimental and lifespan approaches. In E. A. Locke and G. P. Latham (Eds.), New Developments in Goal Setting and Task Performance New York, NY: Psychology Press. West, R. L., & Hastings, E. C. (2011). Self regulation and recall: Growth curve modeling of intervention outcomes for older adults. Psychology and Aging, 26 (4), 803 812. doi:10.1037/a0023784 Wes t, R. L., & Strickland Hughes, C. M. (2015). Memory training for older adults: A review with recommendations for clinicians. In D. Bruno (Ed.), Preservation of Memory New York, NY: Psychology Press. West, R. L., Thorn, R. M., & Bagwell, D. K., (2003). Mem ory performance and beliefs as a function of goal setting and aging. Psychology and Aging, 18 111 125. doi:10.1037/0882 79184.108.40.206 West, R. L., Welch, D. C., & Knabb, P. D. (2002). Gender and aging: Spatial self efficacy and location recall. Basic and Applied Social Psychology, 24 (1), 71 80. doi:10.1207/ S15324834BASP2401_7 West, R. L., Welch, D. C., & Thorn, R. M. (2001). Effects of goal setting and feedback on memory performance and beliefs among older and younger adults. Psychology and Aging, 16 (2), 240 250. doi:10.1037/0882 79220.127.116.11 West, R. L., Welch, D. C., & Yassuda, M. S. (2000). Innovative approaches to memory training for older adults. In R. D. Hill, L. Bckman, & A. S. Neely (Eds.), Cognitive Rehabilitation in Old Age (pp. 81 105). Oxford UK: Oxford University Press. West, R. L., & Yassuda, M. S. (2004). Aging and memory control beliefs: Performance in relation to goal setting and memory self evaluation. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 59 (2), 56 65. doi: 10.1093/geronb/59.2.P56 Wiest, M., Schz, B., & Wurm, S. (2013). Life satisfaction and feeling in control: Indicators of successful aging predict mortality in old age. Journal of Health Psychology 18 (9), 1199 1208. doi:10.1177/1359105312459 099
153 Willis, S. L., Tennstedt, S. L., Marsiske, M., Ball, K., Elias, J., Koepke, K. M., Morris, J. N., Rebok, G. W., Unverzagt, F. W., Stoddard, A. M., & Wright, E. (2006). Long term effects of cognitive training on everday functional outcomes in older adu lts. Journal of the American Medical Association, 296 (23), 2805 2814. doi:10.1001/jama.296.23.2805 Wolinsky, F. D., Vander Weg, M. W., Martin, R., Unverzagt, F. W., Willis, S. L., Marsiske, M., Rebok, G. W., Morris, J. N., Ball, K. K., & Tennstedt, S. L. ( 2009). Does cognitive training improve internal locus of control among older adults? The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 65B (5), 591 598. doi:10.1093/geronb/gbp117 Woolverton, M., Scogin, F., Shackelford, J., Black, S., & Duke, L. (2001). Problem targeted memory training for older adults. Aging, Neuropsychology, and cognitive, 8 (4), 241 255. doi:10.1076/anec.18.104.22.16837 Zehnder, F., Martin, M., Altgassen, M., & Clare, L. (2009). Memory training effects in old age as markers of plasticity: A meta analysis. Restorative Neurology and Neuroscience, 27 (5), 507 520. doi:10.3233/RNN 2009 0491 Zelinski, E. M. (2009). Far transfer in cognitive training of older adults. Restorative Neurology and Neuroscience, 27 (5), 455 471. doi:10.3233/RNN 2009 0495 Zelinski, E. M. (2012). Are strategies necessary to improve memory? Journal of Applied Research in Memory and Cognition, 1 (1). doi:10.1016/j.jarmac.2012.12.004
154 BIOGRAPHICAL SKETCH Carla Marie Strickla nd Hughes received her Bachelor of Science degree in b usiness m anagement from North Carolina State University in August 2011 and her Master of S cience degree of p sychology, with a g erontology certificate, from the University of Florida in August 2014. After completion of her Doc torate of Philosophy in p sychology in August 2017, she will join the Department of Psychology at the University of Pacific in Stockton, California, as a tenure track Assistant Professor. Carla loves spending time with her cat, riding her bicycle, drinking herbal tea, reading fiction, listening to public radio, and playing eurogames.