Skip to main content
NCBI home page
The Journals of Gerontology Series B: Psychological Sciences and Social Sciences logoLink to The Journals of Gerontology Series B: Psychological Sciences and Social Sciences
. 2018 Jun 30;75(4):772–782. doi: 10.1093/geronb/gby081

Daily Control Beliefs and Cognition: The Mediating Role of Physical Activity

Stephanie A Robinson 1,✉, Margie E Lachman 1
Editor: Bob G Knight
PMCID: PMC7328023  PMID: 29982815

Abstract

Objectives

Recent work suggests that physical activity may underlie the positive relationship between control beliefs and cognition. Despite the well-known cognitive benefits, most adults do not engage in enough physical activity, perhaps due to low perceived control. The current study aimed to expand upon past work by investigating these constructs over the short-term by studying the intraindividual variability in daily control beliefs, physical activity, and cognition. We examined whether these constructs were related on a day-to-day basis and if daily physical activity mediated the relationship between control beliefs and cognition.

Method

Over 7 consecutive days, 145 participants (M = 50.54 years) completed daily diaries to measure domain-specific control beliefs, wore an ActiGraph to capture objective physical activity, and were administered 2 tasks each day via phone to measure cognition.

Results

Using multilevel structural equation modeling, our results indicated that on days with higher control beliefs cognition was also higher and this relationship was mediated by one’s level of physical activity.

Discussion

These findings demonstrate the role of physical activity in linking control beliefs and cognition on a daily level using a within-person approach to investigate the dynamic processes in beliefs and cognition.

Keywords: Cognition, Exercise, Mediation Analysis, Perceived Control

Despite the findings that control beliefs decline on average in later adulthood (Lachman & Firth, 2004; Lachman, Neupert, & Agrigoroaei, 2011; Lachman & Weaver, 1998), maintaining one’s sense of control is a well-established predictor of successful aging and low perceived control is related to poor outcomes, such as a decline in cognition (Lachman et al., 2011). Perceived control is linked to enhanced cognition, however only recently have the mechanisms involved in these relationships been explored (Lachman, Agrigoroaei, & Rickenbach, 2015). One proposed mechanism is physical activity (Infurna & Gerstorf, 2013). Recent work supports a longitudinal mediational role of physical activity between adults’ control beliefs and cognition (Infurna & Gerstorf, 2013; Robinson & Lachman, 2016). Control beliefs, physical activity, and cognition are dynamic variables that, not only change across the life span but also vary day-to-day. Utilizing a daily diary paradigm, the present study assessed dynamic variations in control beliefs and how these variations predicted changes in physical activity and cognition. We also assessed if daily physical activity mediated the relationship between daily control beliefs and daily cognition.

Control Beliefs and Cognition

An abundance of evidence suggests individual differences in perceived control are related to pivotal aging outcomes such as cognition (Lachman et al., 2011, 2015). Research has demonstrated that a higher sense of control is associated with better cognition, including better memory (Windsor & Anstey, 2008), greater effective strategy use (Lachman & Andreoletti, 2006), and less cognitive decline (Caplan & Schooler, 2003; Infurna & Gerstorf, 2013; Robinson & Lachman, 2016). Additionally, within-person research has shown that on days in which one perceives more control they are more likely to demonstrate better cognition (Neupert & Allaire, 2012).

Mechanisms

Soederberg-Miller and Lachman (1999) proposed a conceptual model, updated by Robinson and Lachman (2016), concerning the possible mechanisms linking control beliefs and positive health outcomes. Control beliefs are believed to influence outcomes and performance through behavior (e.g., strategy use), physiology (e.g., anxiety, stress), motivation (e.g., effort), or affect (e.g., depression). This model, derived from social cognitive theory (Bandura, 1997), assumes the processes are reciprocal and cyclic such that the outcomes (e.g., memory, physical declines, well-being) may affect one’s control beliefs, self-efficacy, and feelings of mastery, or beliefs about one’s abilities, and/or constraints, which in turn can impact possible behavioral or physiological mediators as well as future outcomes (Bandura, 1997; Soederberg-Miller & Lachman, 1999).

For example, older adults who are experiencing trouble with their memory or physical ability may react with a decreased sense of control in these domains, especially if the difficulties can be attributed to uncontrollable factors (e.g., age, injury). This lowered sense of control can be harmful if associated with increased stress, anxiety, or inactivity (Agrigoroaei & Lachman, 2011). Perceived control and related behaviors seem to be involved in a multidirectional, reciprocal relationship wherein perceived control is both a predictor and outcome of age-related changes such as memory (Soederberg-Miller & Lachman, 1999) and health (Skaff, 2007).

Adaptive Behaviors

The self-regulatory role of adaptive behaviors (e.g., strategy use) has been considered a potential mechanism linking perceived control to healthy aging. The perception of control is tied to better health and cognition especially among older adults (Lachman et al., 2011). Perceived control is likely beneficial for functioning by providing a necessary motivational resource for the development of adaptive and healthy behaviors and strategies used to compensate for limitations or losses (de Frias, Dixon, & Bäckman, 2003; Lachman, 2006).

Physical activity

Conceptual models of perceived control outline that physical activity may be one adaptive behavior through which perceived control is protective against cognitive decline (Lachman et al., 2011; Rodin, 1986; Uchino, 2006). Recent empirical work supports this conceptual model. For example, previous longitudinal studies have demonstrated that perceiving more control is related to engaging in health-promoting behaviors and exhibiting better health profiles (Infurna & Gerstorf, 2012; White, Wójcicki, & McAuley, 2012), which in turn influence cognitive health. Specifically, if one perceives more control, they are more likely to engage in positive health behaviors such as physical activity, and will subsequently capitalize on the cognitive benefits of physical activity.

The relationship between physical activity and cognition is well-established (for review, see Hillman, Erickson, & Kramer, 2008), with evidence from both experimental (e.g., S. Colcombe & Kramer, 2003; S. J. Colcombe et al., 2006; Kramer et al., 1999) and observational paradigms (Whitbourne, Neupert, & Lachman, 2008). Physiological mechanisms to support enhanced cognition as a result of physical activity include increased cerebral blood flow, brain-derived neurotrophic factor (BDNF), changes in neurotransmitter release, and structural changes in the central nervous system (Dishman et al., 2006; Hillman et al., 2008). However, the National Center for Health Statistics reports that only about 20% of adults meet the federal recommended guidelines for aerobic and strength-training activity and that this continues to decrease with age (Clarke, Norris, & Schiller, 2017). The belief that one can exercise, despite constraints and obstacles such as age-related losses, or daily scheduling constraints, is associated with a greater likelihood of engaging in physical activity (Bandura, 1997). Therefore, it is imperative to understand how control beliefs relate to physical activity, as well as cognition.

Intraindividual Variability

There has been a recent trend to investigate the degree to which within-person variation in one variable is associated with within-person variation in another (e.g., Almeida, 2005). Past work has demonstrated that perceived control differs not only between-persons, as a stable individual-difference variable, but also within-persons, as a dynamic variable (Lachman et al., 2015). That is, control beliefs can operate in two ways: generally and situationally. General control beliefs reflect one’s perception of their ability to face demands and obtain desired outcomes in life, whereas situational (e.g., daily) control beliefs reflect this perception in specific situations (Koffer et al., 2017). In fact, stability may be as important as one’s level of control beliefs. Greater variability in control beliefs has predicted cognition more so than one’s level of perceived control (Agrigoroaei, Neupert, & Lachman, 2013). Additionally, a growing number of studies provide evidence of significant and reliable within-person variability in cognition (e.g., Hultsch, MacDonald, & Dixon, 2002; Nesselroade & Salthouse, 2004; Neupert & Allaire, 2012; Sliwinski, Smyth, Hofer, & Stawski, 2006).

Daily Variation

Previous longitudinal investigations (Infurna & Gerstorf, 2013; Robinson & Lachman, 2016) of interindividual differences in intraindividual change in perceived control, physical activity, and cognition provide important examinations of the life-span trajectory of these variables and how they relate to each other. However, these factors not only fluctuate within-persons over the long term but also on a daily basis. Intraindividual variability over the short term is a systematic source of individual differences and has important predictive value (Martin & Hofer, 2004). For example, within-person change in processing speed predicts within-person change in cognition (Sliwinski & Buschke, 2004). Additionally, some studies suggest that older adults show greater intraindividual variability than the young and that such variability is negatively related to performance (Hultsch et al., 2002; Nesselroade & Salthouse, 2004). Apropos the present study, past work has found that days with higher control beliefs are related to days with better cognition (Neupert & Allaire, 2012). Moreover, while most studies have linked physical activity and cognition as a more long-term, cumulative factor, there also exists evidence to support that physical activity offers immediate cognitive benefits as well. For example, Whitbourne and colleagues (2008) found that days with more self-reported daily physical activity were associated with days with better everyday memory. Additionally, Hogan, Mata, and Carstensen (2013) demonstrated immediate cognitive benefits after 15 min of moderate intensity physical activity on that same day.

Current Study

Inherently, cross-sectional and longitudinal studies assume that control beliefs, physical activity, and cognition are relatively stable without daily variation. Therefore, it is important to expand our investigation of interindividual differences (i.e., between-person) in intraindividual change (i.e., within-person) to a short-term scope by examining daily fluctuations. The purpose of the current study was to use a daily diary approach to identify intraindividual variability in control beliefs, physical activity, and cognition, and determine how variation in those constructs is associated within-persons. Examining how daily variation in a mediator such as physical activity influences the relationship between perceived control and cognition promises to illuminate how mediation processes progress on a day-to-day basis. We predicted that those who reported greater perceived control across the week would engage in more physical activity and demonstrate better cognition compared with those low in control. We also predicted that on days in which one reported feeling in more control they would engage in more physical activity and demonstrate better cognition on that day. Finally, we predicted that level of activity would mediate the relationship between control beliefs and cognition on both the between- and within-person levels.

Method

Participants

Participants (n = 145), ranging from 22 to 95 years of age, were recruited mainly from Boston-area communities, and from other regions across the United States. Enrollment was continuous. Participants were recruited using signs in public locations and local newspaper advertisements for the Daily Experiences and Memory Study. Of the 239 who responded to advertisements, 145 were enrolled stratified by age, sex, and education. The remaining 94 were not enrolled due to ineligibility (n = 37), lack of interest (n = 11), or lost contact (n = 46). Our final sample’s mean age was 50.54 (standard deviation [SD] = 19.12). Sixty percent were women, and the mean years of education was 15.59 (slightly less than a bachelor’s degree; SD = 2.47). 70.3% were white, 15.2% were African American, 6.2% reported more than one race, 2.8% did not answer, 2.1% were Asian, and 0.6% reported their race as “other.” 88.3% of participants were not Hispanic, 4.8% were Hispanic or Latin American, and 4.8% did not answer. More descriptive statistics are presented in Table 1.

Table 1.

Descriptive Statistics for All Level 2 Study Variables

Variables N Min Max Mean ICC
Daily control
 Memory 143 1.57 5.00 3.19 .53
 Health 143 1.00 5.00 3.37 .66
 Schedule 143 1.00 5.00 3.31 .33
 Social 143 1.00 5.00 3.46 .52
 Physical activity 143 1.00 5.00 3.38 .55
 Overall 143 1.00 5.00 3.42 .56
Daily MVPA 124 0.42 135.05 36.61 .42
Daily self-report PA 145 1.00 3.10 1.59 .53
Daily cognition
 Daily EM 144 4.80 13.57 9.49 .42
 Daily EF 144 7.29 27.71 15.34 .29
Age 144 22.00 94.00 50.54
Gender 144 1.00 2.00 1.60
Education 144 10.00 20.00 15.18
Functional health 144 1.00 4.00 3.47
Open in a new tab

Note. MVPA = time spent in moderate and vigorous activity; PA = physical activity; ICC = intraclass correlation; EM = episodic memory; EF = executive functioning.

Design and Procedures

The study was approved by the University’s Institutional Review Board. On Day 1 of the study, participants were contacted via telephone and screened for eligibility with the Short Portable Mental Status Questionnaire (SPMSQ; Pfeiffer, 1975). This measure was chosen because it could be easily administered over the phone. It is highly correlated with in-person administration of the measure, and to the Mini-Mental Status Exam (for more details, see Roccaforte, Burke, Bayer, & Wengel, 1994). Participants were deemed eligible if they had no more than two errors. The eligible participants were mailed a package with several background questionnaires, seven daily diaries, and an accelerometer. Participants completed the background questionnaire and mailed it back to the lab in a stamped and addressed envelope. This questionnaire asked about demographics (age, sex, education, functional health, etc.) as well as other measures not relevant to this manuscript. The next day, participants began the first of seven consecutive daily protocols. They were asked to wear the accelerometer on the left hip during all waking hours, and record the exact times they wore the monitor each day. Each night, they completed the daily control beliefs questionnaire, as well as other questionnaires not included in the current analysis. Participants were called each night and reminded to complete the questionnaires and administered two cognitive tasks to measure episodic memory (EM) and executive functioning (EF). If participants did not answer the phone, the researcher waited 5–15 min before calling back. If contact was not made, a voicemail was left asking the participant to call the researcher back. Response rates were high across the study, with an average response rate of 97.2%. Responses were lowest on Day 5 (95.9%) and highest on Day 2 (98.6%). 73.4% of participants completed all 7 days of the phone protocol, and 93% completed 6 or more days of the phone protocol. Participants could earn a total of $100 ($10 for every daily questionnaire mailed back on time and an additional $30 for returning all study materials).

Measures

Covariates

Age was calculated by subtracting the testing date from the participant’s birth date. Gender was measured via self-report and coded as a dummy variable. Education was calculated by recoding a categorical measure into years of education. Functional health was measured with the Physical Functioning subscale from the SF-36 Health Survey (Ware & Sherbourne, 1992). This subscale originally included 10 items that capture the extent to which the participants’ health limits them in different activities (e.g., lifting/carrying groceries, climbing several flights of stairs). Two items asked how much their health limited them in moderate and vigorous activities; as time spent in moderate and vigorous activity was one of our primary variables, we did not include these two items in our composite functional health score. The remaining eight items ranged from 1 (a lot) to 4 (not at all) and were reverse-coded and averaged where a higher score indicated better functional health. Reliability was high (α = .95). Time (e.g., Day 1) was also included as a covariate (Bolger, Davis, & Rafaeli, 2003).

Daily control beliefs

Control beliefs were assessed across six domains (memory, health, physical activity, schedule, social interactions, and things overall) for seven consecutive days. Participants indicated how much control they felt over each domain that day on a scale of 1 (not at all) to 5 (in complete control; e.g., “Today, how much control did you feel over your health?”). This questionnaire was adapted from the domain-specific measure of perceived control (Lachman & Firth, 2004). A latent variable was constructed from each 6 observed items each day. Possible scores for this latent variable ranged from 1 to 5, with a higher number indicating a greater perception of daily control. Reliability was calculated using the 7-day person-mean for each domain and was high (α = .94). Daily reliability for control beliefs ranged from α = .845 on Day 1 to α = .930 on Day 6. The average Cronbach’s α across all 7 days was α = .896.

Daily physical activity

Daily activity was objectively measured with an ActiGraph GT3X+, a small device that measures the acceleration of normal human movements, ignoring high-frequency vibrations associated with the mechanical equipment. Participants were asked to wear the ActiGraph for seven consecutive days and record when the accelerometer was placed and removed to determine wear time, such that only periods the participants indicated he or she was wearing the monitor were included in the analyses. Mean wear time across the 7 days was 12.90 h (SD = 2.78), and ranged from 4 to 22.36 h. Additionally, we used recommended guidelines (Troiano et al., 2008) that required 600 min (10 h) of valid wear-time per day, where 60 min of consecutive zero counts (no movement) was considered nonwear time. Physical activity was operationalized as time spent in moderate-to-vigorous intensity physical activity (MVPA). Using algorithms from Troiano and colleagues (2008), MVPA was calculated in the ActiLife software using “Cut Points” to categorize data into different levels of intensity. Cut point values are based on 60-s epoch lengths.

Cognition

EM and EF tasks were administered via phone for seven consecutive days. EM was assessed using immediate free recall of seven different categorical word lists (Hawkins, Dean, & Pearlson, 2004; Lezak, 1983). Participants were instructed to listen to 15 words that were read out loud at a rate of one second per word and asked to recall as many words as possible. The number of unique words correctly recalled was operationalized as the EM score, with a higher number indicating better EM. To explore the between-person differences of EM, we computed a memory performance score with raw scores averaged across the 7 days. EF was assessed using a category fluency task for seven different categories (Battig & Montague, 1969). Participants were instructed to name as many words from that day’s category as fast as possible. The number of unique and correctly identified words was operationalized as the EF score, with a higher number indicating greater EF. To explore between-person differences, an EF score was computed, where the raw scores were averaged across the 7 days. A latent construct for daily cognition was formed from the observed EM and EF scores.

Data Analysis

Due to the nature of the data with daily occasions nested within-persons, multilevel structural equation modeling (MSEM) was used to analyze the mediation models using Mplus 7.4 for Windows (Muthén & Muthén, 1998–2011). As recommended by Preacher, Zhang, and Zyphur (2011), one advantage of this approach is that it separates the variance of the Level 1 variables (day-level) into between and within components. By doing this, we can consider the fact that the relationships between our variables might differ at both levels. MSEM analyses can reveal information on the effect of daily variations from a person’s mean level of perceived control on that day’s cognition (within-person), the effect of a person’s mean level of perceived control on their mean level of cognition (between-person), and the extent to which these relationships are mediated by physical activity on the within- and between-person levels. The mediated relationship between daily perceived control and cognition is modeled on both the within- and between-person levels and is depicted in Figure 1, which corresponds to a 1-1-1 design where predictor, mediator, and outcome variables are all assessed at Level 1, the day level (Preacher et al., 2011). This model also simultaneously estimates the mediation model on the between-person level (2-2-2).

Figure 1.

Figure 1.

Open in a new tab

Path diagram of the multilevel structural equation mediation model. CB = control beliefs, with 1–6 representing the six domains measured (memory, health, physical activity, social interactions, schedule, things overall). PA = physical activity, C = cognition, with 1–2 representing the two cognitive tests (episodic memory and executive functioning). Figure is based on Preacher and colleagues (2011). All coefficients are standardized estimates. *p < .01; **p < .001.

A latent variable for daily control was constructed with the six observed domains and a latent variable for daily cognition was constructed with two observed variables for EM (word list recall) and EF (category fluency). A measurement model was first constructed to examine model fit of our latent factors (daily perceived control and daily cognition). Parameter constraints were specified by using the MODEL CONSTRAINT command. Indirect effects were specified by using the MODEL INDIRECT command. Confidence intervals were provided. Mplus syntax was adapted from Preacher and colleagues (2011) where we specified a 1-1-1 multilevel model with random intercepts and fixed slopes as specifying all random slopes and intercepts can add unnecessary complications and reduced chance of convergence (Preacher, Zyphur, & Zhang, 2010). Age, gender, education, and functional health were added as Level 2 covariates and day was added as a Level 1 covariate. Mplus is useful in that it uses all data that is available to estimate the model using full information maximum likelihood despite varying number of measurements for each participant, as is common in diary data. Mplus handles missing data by computing the standard errors for the parameter estimates using the observed information matrix (Muthén & Muthén, 1998–2011).

Results

Table 1 contains descriptive statistics for all between-person level variables and the intraclass correlation coefficients (ICC), which represent the ratio of the between-person variation to the total variation. ICC values for control belief domains ranged from 33% to 66%, for schedule and health, respectively. MVPA varied 42% between-persons. ICC values indicated that EM and EF varied 42% and 29% between-persons, respectively. This within-person variance warranted using a multilevel approach. Compliance was high; out of the possible 1,015 observations (145 participants over 7 days), the percentage of missing data ranged from 3.8% (daily control over memory) to 9.1% (daily EF). 25.9% of the MVPA daily data were missing, which included days with less than 600 min of valid wear-time.

Table 2 contains correlations of the Level 2 variables. As hypothesized, all perceived control domain items, MVPA, and cognitive scores were positively correlated with each other. Age was negatively correlated with cognition and MVPA and positively correlated with daily perceived control. Our measurement model was an excellent fit to the data (Confirmatory Fit Index [CFI] = .999 and Root Mean Square Error of Approximation [RMSEA] = .009).

Table 2.

Pearson Bivariate Correlation Coefficients for All Level 2 Study Variables

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1. EM —
2. EF .51** —
3. Cog. composite .96** .74** —
4. Control memory .05 .11 .07 —
5. Control health .19* .18* .21* .68** —
6. Control PA .19* .14 .19* .60** .85** —
7. Control sched. .06 .10 .08 .64** .68** .76** —
8. Control social .15 .15 .17* .65** .67** .73** .82** —
9. Control overall .15 .14 .16 .71** .81** .80** .85** .85** —
10. Control composite .15 .15 .17* .80** .89** .90** .90** .89** .94** —
11. MVPA .21* .26** .26** .05 .23* .21* .04 .03 .10 .13 —
12. Self- report PA .20* .18* .22** .06 .18* .28** .19* .22** .20* .22* .34** —
13. Age −.11 −.29** −.19* .21* .03 .09 .21* .10 .05 .13 −.42** −.12 —
14. Sex −.04 −.05 −.05 −.19* −.24** −.27** −.21* −.11 −.18* −.23** −.34** −.20* .09 —
15. Education .34** .38** .39** .12 .23** .28** .17* .13 .13 .21* .10 .14 .11 −.09 —
16. Health −.03 −.04 −.04 .02 .01 .03 −.04 −.00 −.01 .00 −.02 −.16 .06 .09 .01 —
Open in a new tab

Note. EM = episodic memory; EF = executive functioning; Cog. = cognition; PA = physical activity; Sched. = schedule; MVPA = time spent in moderate and vigorous activity; ICC = intraclass correlation. Day-level data were averaged over 7 days.

*p < .05. **p < .01.

MSEM was used to test the potential mediational effects. Accordingly, this model examined the indirect effect between daily perceived control and daily cognition through daily MVPA at both Level 2 (between-persons) and Level 1(within-persons). Standardized estimates, p-values, and 95% confidence intervals for all direct and indirect effects are provided in Table 3. Contrary to our predictions, there were no significant Level 2 relationships between control beliefs and MVPA, or between MVPA and cognition. Additionally, there was no Level 2 between-person indirect effect of MVPA in the relationship between control beliefs and cognition.

Table 3.

Standardized Estimates, Standard Errors, and 95% Confidence Intervals for the Multilevel Structural Equation Model

Est. SE p 95% CI
Level 1: Within-persons
 Daily control
  Memory 0.65 0.04 <.001 0.571 to 0.724
  Health 0.77 0.03 <.001 0.751 to 0.831
  Physical activity 0.79 0.03 <.001 0.737 to 0.844
  Schedule 0.78 0.02 <.001 0.731 to 0.823
  Social 0.78 0.03 <.001 0.722 to 0.830
  Overall 0.90 0.02 <.001 0.862 to 0.928
 Daily cognition
  EM 0.46 0.08 <.001 0.298 to 0.624
  EF 0.53 0.10 <.001 0.332 to 0.733
 Control → cognition 0.13 0.10 .176 −0.060 to 0.328
 MVPA → cognition 0.33 0.09 <.001 0.158 to 0.495
 Day → cognition −0.01 0.57 .850 −0.123 to 0.101
 Control → MVPA 0.18 0.07 .009 0.044 to 0.309
 Indirect effect 0.25 0.12 .046 0.005 to 0.485
Variance
 Daily control
  Memory 0.58 0.51 <.001 0.482 to 0.680
  Health 0.40 0.05 <.001 0.313 to 0.492
  Physical activity 0.38 0.04 <.001 0.290 to 0.460
  Schedule 0.40 0.04 <.001 0.325 to 0.468
  Social 0.40 0.04 <.001 0.314 to 0.481
  Overall 0.20 0.03 <.001 0.140 to 0.259
 MVPA 0.97 0.02 <.001 0.922 to 1.016
 Daily cognition 0.86 0.06 <.001 0.737 to 0.982
  EM 0.79 0.13 <.001 0.612 to 0.972
  EF 0.86 0.06 <.001 0.737 to 0.972
Level 2: Between-persons
 Daily control
  Memory 0.74 0.05 <.001 0.645 to 0.936
  Health 0.84 0.04 <.001 0.770 to 0.909
  Physical activity 0.86 0.03 <.001 0.792 to 0.918
  Schedule 0.88 0.02 <.001 0.841 to 0.931
  Social 0.88 0.03 <.001 0.823 to 0.940
  Overall 0.96 0.01 <.001 0.929 to 0.984
 Daily cognition
  EM 0.63 0.07 <.001 0.487 to 0.769
  EF 0.78 0.09 <.001 0.612 to 0.953
 Control → cognition 0.09 0.11 .454 −0.138 to 0.308
 MVPA → cognition 0.14 0.17 .389 −0.182 to 0.468
 Age → cognition −0.24 0.11 .037 −0.462 to −0.014
 Sex → cognition 0.09 0.09 .323 −0.088 to 0.267
 Education → cognition 0.41 0.11 <.001 0.199 to 0.617
 Health → cognition 0.24 0.14 .094 −0.041 to 0.521
 Control → MVPA 0.13 0.10 .172 −0.058 to 0.326
 Indirect effect 0.08 0.09 .539 −0.174 to 0.332
 Variance
 Daily control
  Memory 0.45 0.07 <.001 0.310 to 0.593
  Health 0.30 0.06 <.001 0.179 to 0.412
  Physical activity 0.98 0.03 <.001 0.161 to 0.377
  Schedule 0.22 0.04 <.001 0.135 to 0.294
  Social 0.22 0.05 <.001 0.120 to 0.325
  Overall 0.09 0.03 .002 0.032 to 0.325
 MVPA 0.98 0.03 <.001 0.931 to 1.034
 Daily cognition 0.60 0.09 <.001 0.427 to 0.782
  EM 0.61 0.09 <.001 0.428 to 0.783
  EF 0.39 0.14 .004 0.120 to 0.654
Open in a new tab

Note. EM = episodic memory; EF = executive functioning; MVPA = time spent in moderate and vigorous activity; Est. = standardized estimates; SE = standard error; CI = confidence interval. Indirect effect estimates represent unstandardized values.

As predicted, Level 1 within-person analyses demonstrated that on days in which participants reported higher control beliefs, they engaged in more MVPA (B = .18, standard error [SE] = 0.07, p = .009), and on days in which participants engaged in more MVPA they also demonstrated better cognition (B = .33, SE = 0.09, p < .001). There was no significant relationship between days with more perceived control and days with better cognition; however, there was a significant Level 1 indirect effect between daily perceived control and cognition mediated by MVPA, B = .25, SE = 0.12, 95% CI = 0.005–0.485. That is, on days in which one perceived more control compared with their average, they participated in more active minutes of moderate and vigorous intensity and demonstrated better cognitive functioning. The percent mediated was calculated to reflect the proportion of the indirect effect (0.245) to the total effect (0.812). This was 30.2%; therefore approximately 30% of the relationship between daily control and daily cognition can be explained by daily MVPA (Bolger & Laurenceau, 2013).

Due to the design, strong conclusions concerning directionality cannot be made. However, the reversed MSEM model was tested, where daily cognition predicted control beliefs. Between-persons, Level 2 cognition significantly predicted MVPA (p < .001). No other direct or the indirect effects in this model were significant. On the within-person level (Level 1), there was a trend to a significant relationship between MVPA and perceived control (p = .083) and a significant relationship between cognition and MVPA (p = .005). In this model, days with better cognition did not predict days with more perceived control, nor was there a significant indirect effect.

Discussion

The goal of the current study was to examine the relationship between control beliefs cognition, whether this relationship was mediated by physical activity. With a 7-day diary study, we were able to use MSEM to estimate this mediation from both a within (daily) and between-person (averaged across the week) perspective. The results supported our hypotheses on the within-person level but not on the between-person level. That is, on days in which participants reported feeling in more control they also demonstrated better cognition through that day’s level of physical activity. The between-person mediation of physical activity, however, was not significant. Those who perceived more control across the week compared with others did not demonstrate better average cognition, and this relationship was not mediated by average level of activity.

Congruent with previous work, our results show considerable intraindividual variability in daily control beliefs (Eizenman, Nesselroade, Featherman, & Rowe, 1997; Neupert & Allaire, 2012), emphasizing the value of considering control beliefs as an impressionable characteristic susceptible to changes across varying contexts and domains. The present study extends previous work to domain-specific control beliefs and begins to examine the mechanisms linking daily perceived control and cognition.

There was a significant within-person relationship between daily control and objective physical activity, where on days in which participants reported feeling in more control they engaged in more physical activity. Consistent with past work (Whitbourne et al., 2008), there was a significant within-person direct effect, where days with more physical activity were associated with days with better cognition. This study expands upon the work of Whitbourne and colleagues (2008) with objective assessments of both physical activity and cognition. While there was a significant within-person indirect effect, we were unable to replicate Neupert and Allaire’s (2012) work demonstrating a within-person direct effect between days with greater perceived control and days with better cognition. This inconsistency can perhaps be explained by the differences in cognitive tasks. Whereas Neupert and Allaire (2012) found this association specifically for inductive reasoning (and memory in some participants with high average control beliefs), the present study examined EM and EF. Our findings are consistent with Neupert and Allaire (2012) on the between-person level in that neither study found a between-person relationship between average daily control and average daily cognition, highlighting the importance of examining within-person processes as the pattern of results varies across units of analysis (Sliwinski, Hofer, & Hall, 2003).

Participants’ domain-specific perceived control increased with age, while measures of physical activity and cognition decreased with age. These relationships with age are noteworthy given the existing research that links perceived control with better cognition and physical activity (Lachman et al., 2011). Previous work has demonstrated that control beliefs can increase with age in some domains (e.g., work, marriage, finances) and decrease with age in others (e.g., children and sex life), although this has not been examined at the daily level (Lachman & Weaver, 1998). We found that across all domains, daily control beliefs increased with age. Interestingly, general control beliefs, which were examined in the background questionnaire, were negatively related with age, consistent with previous research. It is possible that older adults are better able to take control over their daily lives, and can do so given their vast experience in selecting daily tasks that are within their control. As these constructs and relationships can vary with age, moderated mediation was explored, both linearly, with age as a continuous variable, and categorically, including three age groups (young, middle-aged, and older adults). Age did not significantly moderate the mediational model in either case, which is consistent with other work that has investigated physical activity as a mediator of control beliefs and cognition (Infurna & Gerstorf, 2013; Robinson & Lachman, 2016).

Limitations

One limitation of this study is the use of single measures for the two cognitive abilities. A more multivariate assessment of cognition could provide a better measurement and result in more robust findings. While our reversed model did not yield a significant indirect effect, providing evidence to support our predicted directionality within these relationships, we cannot conclusively state that control beliefs influence cognition and not the other way around. This limitation stems from the design of the study, where all of our measures were assessed on the same day, and thus, we are unable to determine the temporal order in which our participants self-reported their control beliefs. For example, if they were administered the cognitive tasks before filling out the control beliefs questionnaire, their perception of their performance may have influenced their answers. The conceptual model (Lachman, 2006; Robinson & Lachman, 2016; Soederberg-Miller & Lachman, 1999) guiding this work suggests that there is an ongoing reciprocal process, and this may be reflected in the results of the reverse model. Indeed, those who have higher cognitive functioning may, in turn, exercise more and maintain greater perceived control.

Another limitation of this study relates to the varying times of assessment for each participant. Data were collected across the week and start dates and time of day varied by participant. Perceived control, physical activity, and cognition can all be influenced by these temporal factors. For example, there is some evidence to suggest that cognition can vary based on time of day and the individual’s circadian rhythms (Schmidt, Collette, Cajochen, & Peigneux, 2007). Future work should continue to explore this topic and consider including a longer timeframe.

Another limitation is the scale of the control measure; with a response range of 5, it is conceivable that it does not capture as much variation as would be possible with a larger range. Additionally, it is important to consider that this is a subjective measure and could be interpreted differently across participants. Finally, the assumption of independence of residuals between our mediator and outcome measure can be seen as a limitation in our model. However, this assumption is reasonable given the model is recursive, estimating the relationship between the mediator and outcome variable while accounting for covariates, and the model demonstrated good fit indices.

Implications and Future Directions

The present study is unique in its short-term, within-person approach to examining physical activity as a mechanism in the relationship between control beliefs and cognition. Recent empirical work has demonstrated that physical activity is a mechanism in this relationship from a longitudinal perspective (Infurna & Gerstorf, 2013; Robinson & Lachman, 2016), and this study demonstrates that the role of physical activity between control beliefs and cognition is also important on a day-to-day basis. By examining this mediational relationship both as the average across the week and the day-to-day fluctuation both the between- and within-person variation is considered. Within-person variability in control beliefs has been examined for longer-term longitudinal change (Infurna & Gerstorf, 2013) and daily fluctuations (Neupert & Allaire, 2012). However, within-person variation may operate at even more frequent intervals. Therefore, future work could benefit from examining this research question using other time frames (e.g., experience sampling) to capture these temporal variations.

Although there is strong support for the relationship between physical activity and cognition, there are other possible outcomes of physical activity such as positive affect or functional health that could be considered in future research. Another consideration is that in mediation models although directionality can be examined, it is difficult to confirm without an experimental paradigm. Future work should strive to design studies where control beliefs are manipulated to establish directionality and causality. Future research should also investigate how control beliefs are related to physical activity. While the relationship between physical activity and cognition, as well as other health outcomes, is well-documented, a substantial proportion of adults still fail to meet the recommended guidelines for exercise. Therefore, continued work on how control beliefs relate to physical activity and cognition is essential. An interesting finding from this study is the specific item concerning one’s perceived control over their daily schedule. This item demonstrated the most within-person variability compared with the other five items at about 67%, suggesting that participants fluctuate more so in this domain than others and may be particularly important for engaging in physical activity. Perceived time or schedule constraints may play an important role in one’s engagement in physical activity. Future work should expand upon this, and eventually target interventions to promote control beliefs specifically over one’s schedule and time as a means to increase physical activity.

Conclusion

This study is an important step in understanding the mechanisms that link control beliefs to optimal health outcomes typically associated with successful aging, such as good cognitive performance. The objective of this study was to expand upon and further illuminate the nature of the relationships between control beliefs, physical activity, and cognition on a day-to-day basis. While between-person effects are more likely to be seen over the long run (i.e., those who have higher sense of control are more likely to exercise and show less decline in cognition over a longer time frame; Infurna & Gerstorf, 2013; Robinson & Lachman, 2016), the current study specifically examined short-term fluctuations at the daily level and the within-person relationships are more apparent. With this empirical work, we can strengthen previous theoretical models regarding the benefits of having a higher sense of control and expand this model to a within-person framework. Through our diary paradigm, we can identify the importance of control beliefs for engaging in health behaviors and for exhibiting enhanced health outcomes on a daily level. These empirical findings offer a significant advancement to elucidate the mechanisms that are involved in the relationship between one’s personal beliefs and their health and can be used in the future development of evidence-based interventions to enhance important age-related outcomes such as physical activity and cognition.

Funding

This work was supported by the National Institute on Aging (RO1 AG 17920, P30 AG048785, and 5T32AG000204).

Conflict of Interest

None reported.

Stephanie A. Robinson is now at the Department of Veterans Affairs Office of Academic Affiliation Advanced Fellowship Program in Health Services Research, the Center for Healthcare Organization and Implementation Research (CHOIR), Bedford Veterans Affairs Medical Center.

References

  1. Agrigoroaei S. & Lachman M. E (2011). Cognitive functioning in midlife and old age: Combined effects of psychosocial and behavioral factors. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 66(Suppl. 1, i130–i140. doi: 10.1093/geronb/gbr017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. 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 memory. GeroPsych, 26, 45–49. doi: 10.1024/1662-9647/a000078 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Almeida D. M. (2005). Resilience and vulnerability to daily stressors assessed via diary methods. Current Directions in Psychological Science, 14, 64–68. doi: 10.1111/j.0963-7214.2005.00336.x [DOI] [Google Scholar]
  4. Bandura A. (1997). Self-efficacy: The exercise of control. New York: Freeman. [Google Scholar]
  5. Battig W. F., & Montague W. E (1969). Category norms of verbal items in 56 categories: A replication and extension of the Connecticut category norms. Journal of Experimental Psychology, 80(3, Pt. 2), 1–46. doi: 10.1037/h0027577 [DOI] [Google Scholar]
  6. Bolger N. Davis A. & Rafaeli E (2003). Diary methods: Capturing life as it is lived. Annual Review of Psychology, 54, 579–616. doi: 10.1146/annurev.psych.54.101601.145030 [DOI] [PubMed] [Google Scholar]
  7. Bolger N. & Laurenceau J (2013). Within-subject mediation analysis. In Intensive longitudinal methods: An introduction to diary and experience sampling research (pp. 177–195). New York, NY: Guilford Press. [Google Scholar]
  8. Caplan L. J. & Schooler C (2003). The roles of fatalism, self-confidence, and intellectual resources in the disablement process in older adults. Psychology and Aging, 18, 551–561. doi: 10.1037/0882-7974.18.3.551 [DOI] [PubMed] [Google Scholar]
  9. Clarke T. C., Norris T., & Schiller J. S (2017). Early release of selected estimates based on data from the 2016 National Health Interview Survey. Centers for Disease Control. Retrieved from https://www.cdc.gov/nchs/data/nhis/earlyrelease/earlyrelease201705.pdf [Google Scholar]
  10. Colcombe S. J., Erickson K. I., Scalf P. E., Kim J. S., Prakash R., McAuley E.,…Kramer A. F (2006). Aerobic exercise training increases brain volume in aging humans. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences, 61, 1166–1170. doi: 10.1093/gerona/61.11.1166 [DOI] [PubMed] [Google Scholar]
  11. Colcombe S. & Kramer A. F (2003). Fitness effects on the cognitive function of older adults: A meta-analytic study. Psychological Science, 14, 125–130. doi: 10.1111/1467-9280.t01-1-01430 [DOI] [PubMed] [Google Scholar]
  12. Dishman R. K., Berthoud H. R., Booth F. W., Cotman C. W., Edgerton V. R., Fleshner M. R.,…, Zigmond M. J. (2006). Neurobiology of exercise. Obesity (Silver Spring, Md.), 14, 345–356. doi: 10.1038/oby.2006.46 [DOI] [PubMed] [Google Scholar]
  13. Eizenman D. R. Nesselroade J. R. Featherman D. L. & Rowe J. W (1997). Intraindividual variability in perceived control in an older sample: The MacArthur successful aging studies. Psychology and Aging, 12, 489–502. doi:10.1037/0882-7974.12.3.489 [DOI] [PubMed] [Google Scholar]
  14. de Frias C. M. Dixon R. A. & Bäckman L (2003). Use of memory compensation strategies is related to psychosocial and health indicators. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 58, 12–22. [DOI] [PubMed] [Google Scholar]
  15. Hawkins K. A. Dean D. & Pearlson G. D (2004). Alternative forms of the rey auditory verbal learning test: A review. Behavioural Neurology, 15, 99–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hillman C. H. Erickson K. I. & Kramer A. F (2008). Be smart, exercise your heart: Exercise effects on brain and cognition. Nature Reviews. Neuroscience, 9, 58–65. doi: 10.1038/nrn2298 [DOI] [PubMed] [Google Scholar]
  17. Hogan C. L. Mata J. & Carstensen L. L (2013). Exercise holds immediate benefits for affect and cognition in younger and older adults. Psychology and Aging, 28, 587–594. doi: 10.1037/a0032634 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hultsch D. F. MacDonald S. W. & Dixon R. A (2002). Variability in reaction time performance of younger and older adults. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 57, 101–115. doi:10.1093/geronb/57.2.P101 [DOI] [PubMed] [Google Scholar]
  19. Infurna F. J., & Gerstorf D (2012). Perceived control relates to better functional health and lower cardio-metabolic risk: The mediating role of physical activity. Health Psychology, 33, 85–94. doi: 10.1037/a0030208 [DOI] [PubMed] [Google Scholar]
  20. Infurna F. J. & Gerstorf D (2013). Linking perceived control, physical activity, and biological health to memory change. Psychology and Aging, 28, 1147–1163. doi: 10.1037/a0033327 [DOI] [PubMed] [Google Scholar]
  21. Koffer R., Drewelies J., Almeida D. M., Conroy D. E., Pincus A. L., Gerstorf D., & Ram N (2017). The role of general and daily control beliefs for affective stressor-reactivity across adulthood and old age. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 73, 1511–1538. doi: 10.1093/geronb/gbx055 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kramer A. F., Hahn S., Cohen N. J., Banich M. T., McAuley E., Harrison C. R.,…, Colcombe A. (1999). Ageing, fitness and neurocognitive function. Nature, 400, 418–419. doi: 10.1038/22682 [DOI] [PubMed] [Google Scholar]
  23. Lachman M. E. (2006). Perceived control over aging-related declines. Current Directions in Psychological Science, 15, 282–286. doi: 10.1111/j.1467-8721.2006.00453.x [DOI] [Google Scholar]
  24. Lachman M. E., Agrigoroaei S., & Rickenbach E. H (2015). Making sense of control: Change and consequences. In Scott R. A. & Kosslyn S. M. (Eds.), Emerging trends in the social and behavioral sciences (pp. 1–16). Hoboken, NJ: John Wiley & Sons. doi: 10.1002/9781118900772.etrds0209 [DOI] [Google Scholar]
  25. Lachman M. E. & Andreoletti C (2006). Strategy use mediates the relationship between control beliefs and memory performance for middle-aged and older adults. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 61, 88–94. [DOI] [PubMed] [Google Scholar]
  26. Lachman M. E., & Firth K (2004). The adaptive value of feeling in control during midlife. In Brim O., Ryff C., & Kessler R. (Eds.), How healthy are we? A national study of well-being at midlife (pp. 320–349). Chicago, IL: The University of Chicago Press. [Google Scholar]
  27. Lachman M. E., Neupert S. D., & Agrigoroaei S (2011). The relevance of control beliefs for health and aging. In Schaie K. W. & Willis S. L. (Eds.), Handbook of the psychology of aging (7th ed, pp. 175–190). San Diego, CA: Elsevier. doi: 10.1016/B978-0-12-380882-0.00011-5 [DOI] [Google Scholar]
  28. Lachman M. E. & Weaver S. L (1998). Sociodemographic variations in the sense of control by domain: Findings from the MacArthur studies of midlife. Psychology and Aging, 13, 553–562. [DOI] [PubMed] [Google Scholar]
  29. Lezak M. D. (1983). Neuropsychological assessment. New York: Oxford University Press. [Google Scholar]
  30. Martin M. & Hofer S. M (2004). Intraindividual variability, change, and aging: Conceptual and analytical issues. Gerontology, 50, 7–11. doi: 10.1159/000074382 [DOI] [PubMed] [Google Scholar]
  31. Muthén L., & Muthén B (1998–2017). Mplus user’s guide (8th ed). Los Angeles, CA: Retrieved from http://www.statmodel.com/download/usersguide/MplusUserGuideVer_8.pdf [Google Scholar]
  32. Nesselroade J. R. & Salthouse T. A (2004). Methodological and theoretical implications of intraindividual variability in perceptual-motor performance. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 59, 49–55. [DOI] [PubMed] [Google Scholar]
  33. Neupert S. D. & Allaire J. C (2012). I think I can, I think I can: Examining the within-person coupling of control beliefs and cognition in older adults. Psychology and Aging, 27, 742–749. doi: 10.1037/a0026447 [DOI] [PubMed] [Google Scholar]
  34. Pfeiffer E. (1975). A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. Journal of the American Geriatrics Society, 23, 433–441. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1159263 [DOI] [PubMed] [Google Scholar]
  35. Preacher K. J., Zhang Z., & Zyphur M. J (2011). Alternative methods for assessing mediation in multilevel data: The advantages of multilevel SEM. Structural Equation Modeling: A Multidisciplinary Journal, 18, 161–182. doi: 10.1080/10705511.2011.557329 [DOI] [Google Scholar]
  36. Preacher K. J. Zyphur M. J. & Zhang Z (2010). A general multilevel SEM framework for assessing multilevel mediation. Psychological Methods, 15, 209–233. doi: 10.1037/a0020141 [DOI] [PubMed] [Google Scholar]
  37. Robinson S. A., & Lachman M. E (2016). Perceived control and behavior change: A personalized approach. In J. W. Reich & F. J. Infurna (Eds.), Perceived control: Theory, research, and practice in the first 50 years (pp. 201–227). New York, NY: Oxford University Press. [Google Scholar]
  38. Roccaforte W. H. Burke W. J. Bayer B. L. & Wengel S. P (1994). Reliability and validity of the short portable mental status questionnaire administered by telephone. Journal of Geriatric Psychiatry and Neurology, 7, 33–38. [PubMed] [Google Scholar]
  39. Rodin J. (1986). Aging and health: Effects of the sense of control. Science (New York, N.Y.), 233, 1271–1276. [DOI] [PubMed] [Google Scholar]
  40. Schmidt C. Collette F. Cajochen C. & Peigneux P (2007). A time to think: Circadian rhythms in human cognition. Cognitive Neuropsychology, 24, 755–789. doi: 10.1080/02643290701754158 [DOI] [PubMed] [Google Scholar]
  41. Skaff M. M. (2007). Sense of control and health. In Aldwin C. M., Park C. L., & Spiro A. (Eds.), Handbook of health psychology and aging (pp. 186–209). New York, NY: Guilford Press. [Google Scholar]
  42. Sliwinski M., & Buschke H (2004). Modeling intraindividual cognitive change in aging adults: Results from the Einstein aging studies. Aging, Neuropsychology, and Cognition, 11, 196–211. doi: 10.1080/13825580490511080 [DOI] [Google Scholar]
  43. Sliwinski M. J. Hofer S. M. & Hall C (2003). Correlated and coupled cognitive change in older adults with and without preclinical dementia. Psychology and Aging, 18, 672–683. doi: 10.1037/0882-7974.18.4.672 [DOI] [PubMed] [Google Scholar]
  44. Sliwinski M. J. Smyth J. M. Hofer S. M. & Stawski R. S (2006). Intraindividual coupling of daily stress and cognition. Psychology and Aging, 21, 545–557. doi: 10.1037/0882-7974.21.3.545 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Soederberg-Miller L., & Lachman M. E (1999). The sense of control and cognitive aging. In Hess T. M. & Blanchard-Fields F. (Eds.), Social cognition and aging (pp. 17–41). New York: Academic Press. [Google Scholar]
  46. Troiano R. P. Berrigan D. Dodd K. W. Mâsse L. C. Tilert T. & McDowell M (2008). Physical activity in the United States measured by accelerometer. Medicine and Science in Sports and Exercise, 40, 181–188. doi: 10.1249/mss.0b013e31815a51b3 [DOI] [PubMed] [Google Scholar]
  47. Uchino B. N. (2006). Social support and health: A review of physiological processes potentially underlying links to disease outcomes. Journal of Behavioral Medicine, 29, 377–387. doi: 10.1007/s10865-006-9056-5 [DOI] [PubMed] [Google Scholar]
  48. Ware J. E. Jr., & Sherbourne C. D (1992). The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Medical Care, 30, 473–483. [PubMed] [Google Scholar]
  49. Whitbourne S. B. Neupert S. D. & Lachman M. E (2008). Daily physical activity: Relation to everyday memory in adulthood. Journal of Applied Gerontology, 27, 331–349. doi: 10.1177/0733464807312175 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. White S. M. Wójcicki T. R. & McAuley E (2012). Social cognitive influences on physical activity behavior in middle-aged and older adults. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 67, 18–26. doi: 10.1093/geronb/gbr064 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Windsor T. D. & Anstey K. J (2008). A longitudinal investigation of perceived control and cognitive performance in young, midlife and older adults. Neuropsychology, Development, and Cognition. Section B, Aging, Neuropsychology and Cognition, 15, 744–763. doi: 10.1080/13825580802348570 [DOI] [PubMed] [Google Scholar]

Articles from The Journals of Gerontology Series B: Psychological Sciences and Social Sciences are provided here courtesy of Oxford University Press

RESOURCES