Contract Breach and Retirement Among Spanish Adults

Abstract

The present research examines the effects of contract breach on retirement satisfaction. We specifically tested the moderating role of retirement self-efficacy and the mediating role that negative affect may play in influencing the relationships between contract breach and retirement satisfaction, either anticipated or actual. Two empirical studies have been conducted by self-reported questionnaires, with older workers of Spanish descent from Spain—one sample being comprised of workers still in the workforce (Study 1; N = 160) and the other being comprised of recent retirees (Study 2; N = 215). We found that contract breach was positively related to negative affect and negatively related to both anticipated (Study 1) and actual retirement satisfaction (Study 2). Moreover, moderation analyses showed that these effects are strong for participants with low rather than high retirement self-efficacy. Theoretical and practical implications are discussed.

Keywords

contract breach retirement satisfaction affective events theory retirement self-efficacy

A psychological contract is a person’s belief in a mutual agreement that includes obligations between the person and the organization. Unfortunately, employees frequently perceive that the organization has failed to adequately fulfill its obligations, and thus, they develop a perception that their contract with the organization has been breached. Hence, contract breach (CB) has been defined as the individual perception of the extent to which the organization has failed to adequately fulfill its promises and obligations (Robinson & Morrison, 2000). This intense experience has been shown to be an effective predictor of work attitudes and behavior (Zhao, Wayne, Glibkowski, & Bravo, 2007).

More recently, the frontiers of CB have been expanded. Some studies proposed that retirement is a phenomenon belonging to the sphere of work and that attitude toward retirement in older workers and recent retirees will be affected by CB (Bal, De Lange, Jansen, & Van der Velde, 2008, 2013). As already suggested, attitudes toward retirement are formed as a function of the employees’ experiences in the work setting (Newman, Jeon, & Hulin, 2013). Thus, experiences of CB generate negative attitudes because they increase distrust and the perception of unfairness in the treatment the workers receive from their employers. In addition, in the case of older workers, these CB perceptions are linked to broken promises concerning their approaching retirement or to pressures for them to leave their posts prematurely. Such perceptions can lead workers to perceive that their retirement is not completely voluntary but rather, at least in some sense, forced. Nevertheless, till now, the influence of CB on retirement satisfaction has not been empirically explored. Therefore, in this investigation, we intend to deepen this aspect by testing a model of mediators and moderators in the relationship between CB and satisfaction, which will be applied both to older workers and to retirees. This model is based on the arguments that we will present below.

First, overwhelming evidence has been accumulated of the influence of CB on workers’ attitudes and behaviors, including older workers’ attitudes toward their occupational future (Bal & Kooij, 2011). In addition, such attitudes will affect retirees’ long-term well-being and adjustment to the new situation (Bamberger & Bacharach, 2014; Newman et al., 2013). In the retirement literature, both anticipated retirement satisfaction and retirement satisfaction have been conceptualized as adjustment indicators. Hence, given the impact of CB on workers’ attitudes, we propose to examine whether older workers’ anticipated retirement satisfaction and retirees’ retirement satisfaction can be affected by CB.

Second, another aspect that deserves attention is the mechanisms underlying the influence of CB on work attitudes and behaviors, for which we resorted to the affective events theory applied to the psychological contract (Zhao et al., 2007). Empirical evidence has been accumulated to support the mediating role of affect in the relationships between CB and employee attitudes. On the basis of this research, in the present study, we intend to explore whether the negative affect generated by CB can mediate the relationship between CB and anticipated retirement satisfaction in older workers or retirement satisfaction in retirees.

Third, given that people react differently to the influences of negative events, we propose to examine whether CB effects on outcomes will vary as a function of individual differences (Jimmieson, 2000). Empirical evidence supports the moderating role of self-efficacy in the process of occupational stress, and specifically, with regard to retirement, diverse studies have explored the relevance of retirement self-efficacy as an important predictor of well-being and satisfaction in older workers (Harper, 2005). Summing up, this investigation attempts to further our comprehension of the long-term influence of CB on workers’ attitudes, not only within the framework of their jobs but also with reference to their retirement. Moreover, we propose to analyze the mediator and moderator mechanisms of such influence in order to benefit employees, thereby contributing to the development of protective personal resources to cope with the impact of CB on their present and future attitudes.

CB and (Anticipated) Retirement Satisfaction

Research on CB found that inadequate fulfillment of promises made by organizations can be classified according to three dimensions, which in turn constitutes economic (monetary resources not provided by the organization), developmental (advancement and standing in the organization), and socioemotional CB (support and socioemotional concerns; Bal, Jansen, Van der Velde, de Lange, & Rousseau, 2010). CB experiences in the work setting should be linked to broken promises concerning the approaching retirement, and they could decrease voluntariness, which has been in turn associated with more positive mid- and long-term outcomes (Newman et al., 2013). For this reason, in the present research, we expect that CB will be negatively related to anticipated retirement satisfaction, understood as a specific part of job satisfaction that is especially relevant to older workers (Study 1). In the same sense, we propose that CB will be negatively related to retirees’ retirement satisfaction (Study 2).

The Mediating Role of Negative Affect

The reviews and empirical studies have revealed the significance of affects at work, underlining their mediating role in the relationship between working conditions and workers’ attitudes, especially satisfaction (Judge & Ilies, 2004). Drawing from the affective events theory (Weiss & Cropanzano, 1996) applied to the psychological contract, in the present research, we expect that negative affect will mediate the relationship between CB and anticipated retirement satisfaction for older workers or retirement satisfaction for retirees.

Specifically, as mentioned, CB is understood within the framework of the continued relationship between worker and employer. This continued occupational relationship can lead to positive social exchanges that increase employees’ trust and generate positive emotions, thereby favorably influencing anticipated retirement satisfaction and retirement satisfaction. But when CB occurs, negative emotions are triggered, and distrust and anger increase (Robinson & Morrison, 2000; Zhao et al., 2007). Hence, the influence of CB on satisfaction would be based on the generation of intense negative psychological states, which, in turn, lend force to attitudes (Bal & Kooij, 2011). These experiences of organizational breach of promises may act as “affective shocks,” requiring further interpretation and behavioral reaction. As Grandey, Tam, and Brauburger (2002) stated, “When emotional reactions are aggregated over time, they are proposed to influence the overall feelings one has about the job” (p. 32). As a result, CB may increase negative social exchanges and, consequently, in this research, it is predicted that CB will be positively related to negative affect both for older workers (Study 1) and for retirees (Study 2).

The negative relationship between negative affect and anticipated retirement satisfaction for older workers or retirement satisfaction for retirees is based on the premise that the emotions felt in the work setting influence the attitudes toward that specific job and toward retirement, especially when the worker is approaching retirement or has just retired. Van Kleef (2009) proposed the “emotions as social information model,” which explains that affect disambiguate social interaction by providing information about the expresser’s feelings, goals, motives, and intentions. This helps to explain why the emotional reaction following CB—called violation by some authors—is significantly related to a reduction in positive attitudes, such as satisfaction. To sum up, the process through which CB influences the outcomes includes the mediating role of negative affect.

The Moderating Role of Retirement Self-Efficacy

As mentioned, self-efficacy protects people from the harmful effects of stress (Jimmieson, 2000; Schaubroeck & Merritt, 1997). In this sense, the mechanism by which self-efficacy moderates the influences of CB on outcomes could be double. On the one hand, for people with high self-efficacy, CB will have less influence on attitudes, because self-efficacy will compensate for the tendency to blame oneself for the negative event. On the other hand, when workers have low self-efficacy, the harmful impact of CB will be increased by their tendency to perceive the negative event as caused by their own inability to meet the demands of the occupational relationship, thereby reducing their anticipated retirement satisfaction—understood as a specific part of job satisfaction that is especially relevant for older workers—or their retirement satisfaction in the case of retirees.

In contrast, when workers hold firm beliefs about their ability to effectively deal with tasks, including the task of retiring, they will feel confident and optimistic about their future situation, regardless of the negative impact of CB. Their conviction of possessing these skills will benefit their anticipated retirement satisfaction if they are still working or their retirement satisfaction if they have already retired. In other words, employees with low retirement self-efficacy will depend more on the positive experiences provided by their work setting to feel capable of effectively carrying out the transition to retirement. Summing up, if their experiences at the end of their occupational career are frustrating and negative, their anticipated retirement satisfaction if they are still working or their retirement satisfaction if they are retired will drop easily, and all the more so if they lack the psychological skills to offset such adverse experiences.

Moderating effect of retirement self-efficacy in the relationships between negative affect and outcomes

In spite of the fact that the detrimental influence of negative affect on attitudes seems sufficiently supported by empirical research (Judge & Ilies, 2004), researchers have frequently suggested the appropriateness of exploring under what individual determinants this relationship between negative affect and outcomes (anticipated retirement satisfaction for older workers or retirement satisfaction for retirees) is moderate. Specifically, it has been suggested that, although negative affect influences the decrease in satisfaction, this will depend on the extent to which the person has positive psychological resources to offset such negative affect. Therefore, workers with high retirement self-efficacy will have resources to offset the negative effect of retirement self-efficacy on anticipated retirement satisfaction, showing less influence, and the same process will be observed for retirees regarding retirement satisfaction. In other words, through high or low retirement self-efficacy, employers can generate more or less positive conditions so that anticipated retirement satisfaction will develop in the older workers or retirement satisfaction in the retirees. If retirement self-efficacy is a set of resources that help people to perceive that their goals are attainable, this suggests that the harmful influence of negative affect may have less impact on people with high retirement self-efficacy. In contrast, this harmful effect will be increased by low retirement self-efficacy.

The current investigation

Two studies examined the influence of CB on both anticipated retirement satisfaction and retirement satisfaction, considering additional influences from negative affect and retirement self-efficacy.

Study 1

The first study investigated whether negative affect mediates the relationship between CB and anticipated retirement satisfaction and whether retirement self-efficacy moderates the relationship between negative affect and anticipated retirement satisfaction of workers over 55 years. Based on the literature reviewed, we propose the following hypotheses:
Hypothesis 1a: CB will be negatively related to anticipated retirement satisfaction for older workers.

Hypothesis 2a: CB will be positively related to negative affect for older workers.

Hypothesis 3a: Negative affect will mediate the relationships between CB and anticipated retirement satisfaction.

Hypothesis 4a: Retirement self-efficacy moderates the relationship between CB and anticipated retirement satisfaction, such that the relationship is stronger among workers with low rather than high levels of retirement self-efficacy.

Hypothesis 5a: Retirement self-efficacy moderates the relationship between negative affect and anticipated retirement satisfaction, such that the relationship between negative affect and anticipated retirement satisfaction is stronger among participants with low rather than high levels of retirement self-efficacy.

Method

Participants and Procedure

Study 1 was conducted with a final sample of 160 older workers from diverse Spanish organizations located in the capital city of Madrid. Data were collected during 2013. We communicated with 12 medium and small enterprises from the services sector and invited them to participate. Only eight of them agreed, and they provided to the research team access to a total population of 182 workers over 55 years. We contacted the potential participants, explained the objectives and procedures of the study, and assured them of the anonymity and confidentiality of the collected data. At their work sites, we distributed the survey sets containing a cover letter, the informed consent, an envelope for completed questionnaires, and a booklet with the diverse scales of the study. We received 171 completed questionnaires (94% responses). Eleven questionnaires, which have more than 25% of missing data, have been discarded. All of the participants responded after several rounds of follow-up reminders, providing a high response rates. Companies’ sponsorship of the research and the use of work time to complete the questionnaires should be the reasons for these high response rates. Ninety-five percent of the participants were born in Spain and were urban. Mean organizational tenure was 24.4 years (SD = 9.9), and 54% were males. Distribution by sectors in percentages was education (29), health (24), finances (18), public administration (11), and others (19).

Measures

The original questionnaires for CB, negative affect, and retirement self-efficacy were in English, and various experts on retirement translated the items for the Spanish context. A back translation was then donecarried out by a native English speaker, which was compared with the original questionnaires. For the anticipated retirement satisfaction measure, a Spanish version was available.

CB

Employer non fulfillment was measured by indicating the extent to which employees believed their employer had non fulfilled a range of obligations. The instructions for respondents were “To what extent did the organization break the following obligations?” This measure of CB has been long used since 2005 (Coyle-Shapiro & Conway, 2005), and recently, it has been applied to studies of aged workers (Bal et al., 2013; Bal et al., 2010). Coyle-Shapiro and Conway (2005) conducted an exploratory factor analysis and found three dimensions, while subsequent studies supported the reliability of the CB measures with this instrument in several applications among the aged samples. First, in employees aged 40 and more, reliabilities at α = .85 for the economic dimension, with only 5 items at α = .77 for the socioemotional breach (Bal et al., 2013). Second, with employees at a medium age of 54.8 years, the CB dimension showed adequate reliability ranging from α = .90 for economic to α = .84 for socioemotional (Bal et al., 2013). And finally, among postretirement workers (Bal et al., 2010), the α reliabilities for economic dimension was α = .87, for socioemotional α = .78, and for developmental α = .79. All the dimensions of breach identified by previous research have been included in the present study: economic, socioemotional, and developmental psychological contract non fulfillment. Economic breach was measured with 3 items (e.g., “Fair pay for the responsibilities of my job”). Due to the fact that promises and obligations are mutually exchanged by organizations and employees, the list of items in a CB scale is contingent to each relationship. In the present study, following the selection procedure recommended in the CB literature (Bal et al., 2013), we retained only 3 items from the economical dimension. A previous checking, whether psychological contract obligations were indeed part of the perceived promises made by the organizations, has been made by interviews with two union representatives. The following CB dimensions were measured with the full subscales. Socioemotional includes 4 items (e.g., “Freedom to do the job well”) and developmental breach with 3 items (e.g., “Career support and mentoring”). Following the same procedure, union representatives have revised the items and concluded that the 4 socioemotional items and the 3 developmental items should be considered part of the psychological contracts. In our study, reliabilities ranged from α = .79 to α = .78. Scale responses ranged from 1 (not at all) to 5 (to a very great extent).

Negative affect

Based on the affective events theory, we assessed the negative affect on work by using items from the Job Emotion Scale (JES; Fisher, 2000). In this research, only five specific negative emotional states were retained. This decision was based on two requirements: (a) the items represented a variety of negative emotions and (b) they were identified as having essentially the same meaning in English and Spanish. Sample items are “frustrated,” “angry,” and “unhappy,” and the scale responses ranged from 1 (never) to 5 (daily). Fisher (1997) reported a reliability value for the negative emotions subscale of JES of α = .81, with a sample of 116 university students who held jobs. Subscale reliability in our study was α = .84. Factor validity has been reported for the JES author with the global scale of 16 adjectives, obtaining two factors (positive and negative), which accounted for 49% of variance (Fisher, 1997).

Retirement self-efficacy

We used 13 items of the Retirement Self-Efficacy Scale (Harper, 2005), which contained two subscales: activities (8 items) and retirement itself (4 items). Selection criteria for the subscales were based on reliability values in previous studies with aged participants, such as Harper (2005), where αs ranged from .93 to .97 (α = .96 for males, α = .97 females, α = .93 minorities, and α = .97 Caucasians). Response scales ranged from 1 (very little confidence) to 5 (quite a lot of confidence). Example of item from the retirement itself subscale is “plan leisure time.” For the activities subscale, example is “maintain any current skills or knowledge that you want to maintain.” Prior studies with the 13-item version of the scale reported reliabilities of α = .97 (Topa & Alcover, 2015). In this study, for the global scale the reliability was α = .77.

Anticipated retirement satisfaction

As a sensitive measure of specific facets of life near retirement, we used the Spanish adaptation of satisfaction with life subscale of the retirement satisfaction inventory (RSI), which included 11 items (Fernández, Crego, & Alcover, 2011). Floyd et al. (1992) reported reliabilities for each factor in their study with the English version (satisfaction with services, α = .78; satisfaction with health, α = .72; and satisfaction with marriage and family life, α = .57). The Spanish RSI has been previously applied to sample pensioners with Cronbach’s α reliability of .81 (Topa & Alcover, 2015). In our study, the scale reached acceptable reliability values (α = .81). The response scale ranged from 1 (totally unsatisfied) to 5 (totally satisfied).

The composite reliability (CR) has been proposed as an alternative to Cronbach’s α. Its value must be greater than .70 and less than .60, indicating a lack of reliability. In this study, the CR was satisfactory for all the scales (contract breach CR = .75, retirement self-efficacy CR = .92, and anticipated retirement satisfaction CR = .84).

Control variables

We also included individual sociodemographic data in the questionnaire. Participants provided information about their age, seniority in their organizations, and employed time or retired time. Demographic information, such as country of birth and region of living (rural vs. urban), have been included.

Convergent validity represents the common variance between indicators and their construct. Fornell and Larcker (1981) recommended the use of the average variance extracted (AVE) as a criterion. The higher the AVE value, the more representative are the indicators of the construct on which they load. It is suggested that its value must be greater than .50 but 0.4 is also acceptable as a cutoff. Fornell and Larcker said that if the AVE is less than 0.50 but the CR is higher than 0.6, the convergent validity of the construct is still adequate. In this study, the AVE was acceptable (contract breach AVE = .61, retirement self-efficacy AVE = .50, and anticipated retirement satisfaction AVE = .40). To assess discriminant validity among constructs, the AVE square root should be higher than the correlation between constructs. The results in our study show that there is discriminant validity for all the constructs.

Cautions against common method variance

As self-report questionnaires were used to collect the data at the same time from the same participants, common method variance may be a concern. In order to test whether variance in the data can be largely attributed to a single factor, we used the post hoc Harman one-factor analysis (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). In order to apply the post hoc procedure, we examined the results of preliminary exploratory factor analysis which showed that a single factor only accounted for 24% of the variance. Hence, a single factor cannot account for the variance in the data, and we cannot consider the common method variance to be a serious deficiency in this data set. In order to test the hypotheses, we conducted multiple regression analyses (stepwise procedure), entering the control variables in the first model and the predictor ones in the second model.

Results

Hypotheses Testing

Table 1 (Panel A) presents descriptive statistics for Study 1 variables and Pearson’s correlation coefficients. CB was significantly and positively related to negative affect (r = .48, p < .01) and negatively related to anticipated retirement satisfaction (r = −.31, p < .01). Negative affect correlated significantly and negatively with anticipated retirement satisfaction (r = −.37, p < .01). Retirement self-efficacy correlated significantly and negatively with CB (r = −.63, p < .01) and negative affect (r = −.63, p < .01) and positively with anticipated retirement satisfaction (r = .34, p < .01).

Table 1.
Means, Standard Deviations, and Correlations for Studies 1 and 2.

A. Study 1 (aged workers)^a

Variables M SD 1 2 3 4 5 6

1. Age 57.7 3.8 –

2. Job seniority 37 8.5 .37 –

3. CB 2 .78 .08 .09 –

4. NA 1.92 1.06 .18 .12 .48 –

5. ARS 3.46 .52 −.16* −.08 −.31 −.37 –

6. RSE 4.1 .67 −.18* −.20* −.63 −.63 .34 –

B. Study 2 (retirees)^b

Variables M SD 1 2 3 4 5 6

1. Age 65.6 6.4 —

2. Retirement time (years) 1.5 892 .76 —

3. CB 2.71 1.08 −.02 .00 —

4. NA 2.63 .57 −.04 −.07 .32 —

5. RS 3.50 1.36 .12 −.00 −.18 −.15 —

6. RSE 3.90 .71 .05 .04 −.15 −.28 .26 —

Note. CB = contract breach; NA = negative affect; RSE = retirement self-efficacy; ARS = anticipated retirement satisfaction; RS = retirement satisfaction.

^a N = 160. ^b N = 215.

p < .05. p < .01.

Table 2 summarizes the results of regression analysis for testing Hypotheses 1a and 2a. The second step presented in Table 2 indicates that the R ² associated with CB was significant (Step 2, β = −.29, p < .001), showing support for Hypothesis 1a. Table 2 shows the results regarding Hypothesis 2a, where the R ² change associated with CB was significant (Step 2, β = .46, p < .001).

Table 2.
Results of Regression for Study 1 (Hypotheses 1a and 2a) and for Study 2 (Hypotheses 1b and 2b).

Predictors Criterion Variable: ARS Criterion Variable: RS

Study 1 (Hypothesis 1a) Study 2 (Hypothesis 1b)

β β

Step 1 Step 2 Step 1 Step 2

Controls

Age −.16 −.14 .36 .35*

Job seniority −.02 −.00 −.29 −.28

Independent variable

CB −.29* −.19

F 2.2 13.5* 6.5 8.1*

R ² .03 .11 .06 .09

▵R ² .03 .08 .06 .03

Criterion Variable: NA Criterion Variable: NA

Study 1 (Hypothesis 2a) Study 2 (Hypothesis 2b)

β β

Predictors Step 1 Step 2 Step 1 Step 1

Controls

Age .16 .11 −.05 −.04

Job seniority .06 .00 −.03 −.04

Independent variable

CB .46* .32*

F 1.43 21* .60 23.65*

R ² .04 .24 .006 .11

▵R ² .04 .20 .006 .10

Note. CB = contract breach; ARS = anticipated retirement satisfaction; RS = retirement satisfaction; NA = negative affect.

p < .05. p < .01. p < .001.

To test Hypothesis 3a, we used the SPSS (version 22) macro developed by Preacher and Hayes (2008). Due to the fact that we hypothesize that negative affect mediates the effects of CB on anticipated retirement satisfaction, we used the indirect macro with 5,000 bootstrapped resamples (CB as independent variable, negative affect as mediator, and age and job seniority as covariates). The results showed that the relationship between CB and anticipated retirement satisfaction was significantly mediated by negative affect (R ² = .17, p < .01). Moreover, both the path from CB to negative affect (.57, p < .001) and the total effect of CB on anticipated retirement satisfaction (−.18, p < .05) were significant. Specifically, the indirect effect of CB on anticipated retirement satisfaction through negative affect was −.08, and the 95% bias-corrected confidence interval (CI) around the bootstrapped indirect effect did not contain zero (bias-corrected CI = [−.19, −.005]). These results indicate that workers who perceived high CB reported high negative affect, which in turn was negatively related to higher anticipated retirement satisfaction. Thus, our Hypothesis 3a was supported (see Figure 1a).

Figure 1.
(a) Mediated moderation model for Study 1 (aged workers) and (b) mediated moderation model for Study 2 (retirees).

Table 3 summarizes the results regarding Hypotheses 4a and 5a. Regarding Hypothesis 4a, the R ² change associated with the addition of the interaction term was also significant, indicating that the interaction between CB and retirement self-efficacy is significant (Step 4, β = .31, p < .001). Related to Hypothesis 5a, the R ² change associated with the addition of the interaction term was significant (Step 4, β = −.31, p < .01), lending support to Hypothesis 5a, which proposed that retirement self-efficacy moderates the relationship between negative affect and anticipated retirement satisfaction. Figure 2 shows that the pattern of interaction between CB and retirement self-efficacy was consistent with Hypothesis 4a. This figure shows that there is a stronger negative relationship between CB and anticipated retirement satisfaction when retirement self-efficacy is low. Whereas CB was unrelated to anticipated retirement satisfaction when retirement self-efficacy is high, it was negatively related to anticipated retirement satisfaction when retirement self-efficacy is low.

Table 3.
Results of Regression Testing for Study 1 (Hypotheses 4a and 5a) and for Study 2 (Hypotheses 4b and 5b).

Predictors Study 1, Hypothesis 4a Study 2, Hypothesis 4b

Criterion Variable: ARS Criterion Variable: RS

Step 1 β Step 2 β Step 3 β Step 4 β Step 1 β Step 2 β Step 3 β Step 4 β

Controls

Age −.16 −.14 −.12 −.12 .36* .36* .41* .41*

Job seniority −.02 −.00 .02 .05 −.29 −.29 −.34 −.35*

Independent variable

CB −.29*** −.16 −.14 −.14* −.14*

F 2.19 13.5* 6.46 6.46**

R ² .03 .11 .05 .05

▵R ² .03 .08 .05 .05

Moderation

RSE .20* .06 .29* .29*

CB × RSE .31** −.12*

F 3.7* 12.5 19.43* 3.76*

R ² .13 .20 .17 .19

▵R ² .02 .07 .08 .02

Study 1, Hypothesis 5a Study 2, Hypothesis 5b

Predictors Criterion variable: ARS Criterion variable: RS

Step 1β Step 2β Step 3β Step 4β Step 1β Step 1β Step 3β Step 4β

Controls

Age −.15 −.10 −.09 −.08 .36* .36* .42* .42*

Job seniority −.08 −.06 −.03 −.04 −.29 −.29 −.35* −.34*

Independent variable

NA −.35*** −.23 −.17 −.13* −.04 −.05

F 1.58 13.5* 6.46 3.80*

R ² .04 .16 .05 .07

▵R ² .04 .12 .05 .02

Moderation

RSE .19+ .06 .29* .31*

NA × RSE −.31** −.11+

F 2.9* 5.5 19.05* 2.83+

R ² .18 .24 .15 .16

▵R ² .02 .06 .08 .01

Note. CB = contract breach; ARS = anticipated retirement satisfaction; RSE = retirement self-efficacy; NA = negative affect.

+p < .10. p < .05. p < .01. p < .001.

Figure 2.
Moderating effect of retirement self-efficacy (RSE) on contract breach (CB)—anticipated retirement satisfaction left (Study 1, Hypothesis 4a) and moderating effect of RSE on CB—retirement satisfaction right (Study 2, Hypothesis 4b).

To summarize, CB was positively related to anticipated retirement satisfaction. Second, CB was positively related to negative affect. Third, negative affect mediates the effects of CB on anticipated retirement satisfaction. Fourth, retirement self-efficacy negatively moderates the relationship between CB and anticipated retirement satisfaction. Finally, retirement self-efficacy moderates the relationship between negative affect and anticipated retirement satisfaction, such that the relationship between negative affect and anticipated retirement satisfaction is stronger among participants with low rather than high levels of retirement self-efficacy. Subsequently, a detailed general discussion is available, comprising Studies 1 and 2.

Study 2

The second study investigated whether negative affect mediates the relationship between CB and retirement satisfaction and whether retirement self-efficacy moderates the relationships between negative affect and retirement satisfaction of recent retirees. Study 2 expanded this model with a new sample of retirees examining the mediation role of negative affect in the CB-retirement satisfaction relationships, and the moderation of retirement self-efficacy in the negative affect–retirement satisfaction relationships:
Hypothesis 1b: CB will be negatively related to retirement satisfaction for retirees.

Hypothesis 2b: CB will be positively related to negative affect for retirees.

Hypothesis 3b: Negative affect will mediate the relationships between CB and retirement satisfaction for retirees.

Hypothesis 4b: Retirement self-efficacy negatively moderates the relationship between CB and retirement satisfaction for retirees, such that the relationship is stronger among retirees with low rather than high levels of retirement self-efficacy.

Hypothesis 5b: Retirement self-efficacy moderates the relationship between negative affect and retirement satisfaction for retirees, such that the relationship between negative affect and retirement satisfaction is stronger among participants with low rather than high levels of retirement self-efficacy.

Method

Participants and Procedure

Study 2 included retired persons. Workers retired during the 2 previous years were located through the archives of the human resources departments of 50 medium-size logistic firms (less than 100 employees) located in Madrid. The firms were invited to participate in the research project through the collaboration of their human resource managers enrolled in a master’s programs at the university (Universidad Nacional de Educación a Distancia) in Madrid, making this work an exchange for academic credits. The research team contacted the potential participants by e-mail, explained to them the purpose of our study, and invited them to participate in exchange for a lottery reward. The study population constituted 265 workers retired during the 2 previous years, who have been contacted by e-mail. Two hundred and forty-eight survey package sets—cover letter, informed consent, questionnaire, and the return envelope—were sent to those who agreed to participate. Completed questionnaires and informed consent were mailed back to the researchers using the prepaid return envelope. One month later, a reminder was sent out to all participants to return the survey. High response rate was attributed to the round of reminders and the lottery reward. We received 230 questionnaires (93% response rate), but we discarded 15 due to the fact that these had more than 25% of missing data. Therefore, the final sample for Study 2 included 215 retirees. Participants’ mean organizational tenure was 31.2 years (SD = 6.9), whereas the percentage of men was 60.6%. Regarding their duties when they were active, 67% were blue-collar, and of the remaining 33%, 7.8% were directors of their companies.

Measures

Predictor variables for this study were similar to Study 1. Sociodemographic data, CB, negative affect, and retirement self-efficacy were measured as per Study 1. CB subscales’ reliabilities ranged from α = .79 to α = .81. Negative affect reliability was α = .79. Retirement self-efficacy reliability was α = .89. In this study, the CR was satisfactory for all the scales (contract breach CR = .70, retirement self-efficacy CR = .85, and retirement satisfaction CR = .89).

Retirement satisfaction

We used the Spanish adaptation by Fernández, Crego, and Alcover (2011) of satisfaction with life subscale of the RSI, which included 11 items. Cronbach’s α reliability for the original subscale was α = .81 in the Spanish version. In our study, the scale reached acceptable reliability values (α = .83). The response scale ranged from 1 (totally unsatisfied) to 5 (totally satisfied).

Convergent validity has been assessed by the AVE values, which were acceptable (contract breach AVE = .71, retirement self-efficacy AVE = .45, and retirement satisfaction AVE = .49) considered in their relationship with the CR values (Fornell & Larcker, 1981). To assess discriminant validity among constructs, the AVE square root should be higher than the correlation between constructs. In our study, the results show discriminant validity for all the constructs.

Results

The results of preliminary exploratory factor analysis showed that a single factor only accounted for 23.5% of the variance (Podsakoff et al., 2003), supporting that we cannot consider the common method variance to be a serious deficiency in this data set.

Hypotheses Testing

Table 1 (Panel B) presents descriptive statistics for Study 2 variables and Pearson’s correlation coefficients. CB was significantly and positively related to negative affect (r = .32, p < .01) and negatively related to retirement satisfaction (r = −.18, p < .01). Negative affect correlated significantly and negatively with retirement satisfaction (r = −.15, p < .05). Retirement self-efficacy correlated significantly and negatively with CB (r = −.15, p < .05), negative affect (r = −.28, p < .01), and positively with retirement satisfaction (r = .26, p < .01).

Table 2 summarizes the results of regression analysis for testing Hypothesis 1b (above) and Hypothesis 2b (below). The second step presented in Table 2 indicates that the R ² associated with CB was significant (Step 2, β = −.19, p < .01), showing support for Hypothesis 1b. Regarding Hypothesis 2b, the R ² change associated with CB was significant (Step 2, β = .32, p < .001).

Finally, to test Hypothesis 3b, we used the SPSS macro developed by Preacher and Hayes (2008). The results showed that the relationship between CB and retirement satisfaction was significantly mediated by negative affect (R ² = .10, p < .001). Moreover, both the path from CB to negative affect (.32, p < .001) and the total effect of CB on retirement satisfaction (−.19, p < .01) were significant. Specifically, the indirect effect of CB on retirement satisfaction through negative affect was −.03, and the 95% bias-corrected CI around the bootstrapped indirect effect did not contain 0 (bias-corrected CI = [−.08, −.02]). These results indicate that workers who perceived high CB reported high negative affect, which in turn was negatively related to higher retirement satisfaction. Thus, our Hypothesis 3b was supported (see Figure 1b).

Table 3 summarizes the results regarding Hypotheses 4b and 5b. According to Hypothesis 4b, the R ² change associated with the addition of the interaction term was also significant, indicating that the interaction between CB and retirement self-efficacy is significant (Step 4, β = −.12, p < .05). Figure 2 shows that the pattern of the interaction between CB and retirement self-efficacy was consistent with Hypothesis 4b. Thus, the relationship between CB and retirement satisfaction was stronger when retirement self-efficacy was lower rather than high. As shown in Table 3, the R ² change associated with the addition of the interaction term was not significant (Step 4, β = −.11, p < .10), showing only limited support for Hypothesis 5b, which proposed that retirement self-efficacy moderates the relationship between negative affect and retirement satisfaction.

To summarize, CB was positively related to retirement satisfaction. Second, CB was positively related to negative affect. Third, negative affect mediates the effects of CB on retirement satisfaction. Fourth, retirement self-efficacy negatively moderates the relationship between CB and retirement satisfaction. Finally, retirement self-efficacy moderates the relationship between negative affect and retirement satisfaction, such that the relationship between negative affect and retirement satisfaction is stronger among participants with low rather than high levels of retirement self-efficacy.

Discussion

The main aim of this study was to examine the role that older workers’ and retirees’ retirement self-efficacy may play in the relationship between CB and retirement satisfaction. First, the results showed that the impact of CB on anticipated retirement satisfaction for older workers and retirement satisfaction for retirees was moderated by participants’ individual differences in retirement self-efficacy. This relationship was stronger for older workers and retirees with low retirement self-efficacy than for those with high levels of retirement self-efficacy. Second, we analyzed the mediating role of negative emotional reactions to CB in the relationship between CB and retirement satisfaction. The results of this research provided support to the hypotheses, and these findings have both theoretical and practical implications.

Theoretical Implications

The first implication is that older workers’ and retirees’ CB affects them emotionally, mentally, and socially. When coupled with low retirement self-efficacy, it has a strong impact on retirement satisfaction, either actual or anticipated. But our findings also suggest that there is a complementary influence between organizational fulfillment of promises and the psychological resources of employees, due to the fact that retirement satisfaction—either actual or anticipated—strongly decreases when a lack of retirement self-efficacy was complemented with high CB. The second implication is that our results also empirically show that the negative affective reactions experienced by workers play a mediating role in the relationship between CB and retirement satisfaction. This contribution highlights the mediational role of negative affect in the relationships between CB and retirement satisfaction for both older workers and retirees. Contrary to the recent findings reporting that some negative behaviors were direct responses of CB (Wang & Hsieh, 2014), our results showed that attitudinal responses should be mediated by negative affect, supporting the sequence proposed by Zhao, Wayne, Glibkowski, and Bravo (2007). Finally, our results are in line with recent evidence suggesting that a combination of person-based antecedents and work-related factors should impact on retirement decisions and adjustment (Bamberger & Bacharach, 2014).

Limitations of the Present Research

In this sense, some limitations of this work must be mentioned. First, our analyses are based on cross-sectional data and cannot be used to test causal relations. Due to the fact that the mediator effect studied is coherent with causal relations, future studies should use longitudinal designs to establish such relations more solidly. Second, prior studies that have focused on older workers have used participants of other European countries, but no studies were carried out applying the model of psychological contract to older workers in Spain. Despite the fact that our sample proceeds from one country, we think that the processes analyzed are not specific to any concrete occupational setting, so our results are generalizable to people from other regions and cultures.

Implications for Practitioners

Related to the practical implications, we could affirm that empirical research on both older workers’ and retirees’ satisfaction has underlined its importance in people’s subsequent adjustment, and a broad array of predictors of satisfaction have been explored. However, the importance of older workers’ experiences during the last stages of their professional life and the potential impact on their subsequent adjustment has often been ignored. In this investigation, extending the model of psychological contract, we have considered that the CB experienced by older workers can negatively influence their retirement satisfaction. When aged employees experience CB while working, they develop something that may be called distrust or dislike of their employers and subsequently conclude that their retirement lives that partly depend on employers’ actions (pension provisions, etc.) are also not likely to be satisfying. In this sense, the study can serve as an advance notice to the directors and managers of human resources of firms to attend to fulfilling the psychological contract of older workers. On the one hand, the impact of CB is negative and long-lasting in older people. On the other hand, the internal social responsibility of the firms’ demands they attend to all the workers’ rights, but particularly, to those of the most vulnerable. Currently, older employees may be one of these disadvantaged older groups (Rauschenbach, Göritz, & Hertel, 2012). Although individuals’ perceptions of CB may not be directly controllable, in contrast, organizational events that trigger such perceptions are subject to intervention. Moreover, if employees’ psychological resources, such as retirement self-efficacy, protect them from the adverse effect of CB on subsequent attitudes, retirement self-efficacy could be considered as a source of competitive advantage for teams and organizations. As the empirical research suggested that specific self-efficacy could be developed by training interventions, the possibility of reducing the negative impact of CB on retirement satisfaction should be taken into account.

Considering that CB is often proximal to the experience of forced retirement for older workers, the present research provides additional evidence of the negative influence of involuntary retirement on mid- and long-term outcomes. Again, our data can warn human resource managers about the appropriateness of proposing programs to prepare older workers for retirement. Such programs would allow them to progressively adjust their expectations to the new situation and foresee in advance how to satisfy their needs as retirees.

Lastly, additional managerial implications can be derived from our findings of the mediating role of negative affect on retirement attitudes. This evidence is relevant for gerontological practitioners because it shows that negative emotions during the retirement transition are plausible, especially due to its difficult or stressful characteristics. Our research illustrates that older workers may feel sad when they realize that they will lose their close colleagues because of their retirement. On the basis of this evidence, practitioners could design more efficient intervention programs if they take into account older people’s affective experiences, especially because such affective experiences have been shown to exert an important influence on older people’s life satisfaction. Specifically, we recommend that future research identify specific events that are hypothesized as causing positive affective reactions and reducing negative experiences. As retirement is a very relevant transition for people, an increase in uncertainty is experienced by older workers. Therefore, as Robinson and Morrison (2000) suggested, the individual need for information increases. If this personal need is not satisfied by the organization, employees’ perceptions of CB can be triggered, provoking negative emotional reactions that will have a long-term impact on people’s well-being. Summing up, retirement can be seen as a distancing of people from their prior working experiences, which allows them to enjoy a new stage. However, our results show that intense emotions are associated with the fact of retiring, and they may persist even 2 years after, influencing long-term well-being.

A. Study 1 (aged workers)^a
Variables	M	SD	1	2	3	4	5	6
1. Age	57.7	3.8	–
2. Job seniority	37	8.5	.37**	–
3. CB	2	.78	.08	.09	–
4. NA	1.92	1.06	.18	.12	.48**	–
5. ARS	3.46	.52	−.16*	−.08	−.31**	−.37**	–
6. RSE	4.1	.67	−.18*	−.20*	−.63**	−.63**	.34**	–
B. Study 2 (retirees)^b
Variables	M	SD	1	2	3	4	5	6
1. Age	65.6	6.4	—
2. Retirement time (years)	1.5	892	.76**	—
3. CB	2.71	1.08	−.02	.00	—
4. NA	2.63	.57	−.04	−.07	.32**	—
5. RS	3.50	1.36	.12	−.00	−.18**	−.15**	—
6. RSE	3.90	.71	.05	.04	−.15**	−.28**	.26**	—

Predictors	Criterion Variable: ARS	Criterion Variable: RS
Controls
Age	−.16	−.14	.36***	.35***
Job seniority	−.02	−.00	−.29**	−.28**
Independent variable
CB		−.29***		−.19**
F	2.2	13.5***	6.5**	8.1***
R ²	.03	.11	.06	.09
▵R ²	.03	.08	.06	.03
	Criterion Variable: NA	Criterion Variable: NA
Study 1 (Hypothesis 2a)	Study 2 (Hypothesis 2b)
β	β
Predictors	Step 1	Step 2	Step 1	Step 1
Controls
Age	.16	.11	−.05	−.04
Job seniority	.06	.00	−.03	−.04
Independent variable
CB		.46***		.32***
F	1.43	21***	.60	23.65***
R ²	.04	.24	.006	.11
▵R ²	.04	.20	.006	.10

Predictors	Study 1, Hypothesis 4a	Study 2, Hypothesis 4b
	Step 1 β	Step 2 β	Step 3 β	Step 4 β	Step 1 β	Step 2 β	Step 3 β	Step 4 β
Controls
Age	−.16	−.14	−.12	−.12	.36***	.36***	.41***	.41***
Job seniority	−.02	−.00	.02	.05	−.29**	−.29**	−.34**	−.35***
Independent variable
CB		−.29***	−.16	−.14			−.14*	−.14*
F	2.19	13.5***			6.46**	6.46**
R ²	.03	.11			.05	.05
▵R ²	.03	.08			.05	.05
Moderation
RSE			.20*	.06			.29***	.29***
CB × RSE				.31**				−.12*
F			3.7*	12.5**			19.43***	3.76*
R ²			.13	.20			.17	.19
▵R ²			.02	.07			.08	.02
	Study 1, Hypothesis 5a	Study 2, Hypothesis 5b
Predictors	Criterion variable: ARS	Criterion variable: RS
	Step 1β	Step 2β	Step 3β	Step 4β	Step 1β	Step 1β	Step 3β	Step 4β
Controls
Age	−.15	−.10	−.09	−.08	.36***	.36***	.42***	.42***
Job seniority	−.08	−.06	−.03	−.04	−.29**	−.29**	−.35***	−.34***
Independent variable
NA		−.35***	−.23	−.17		−.13*	−.04	−.05
F	1.58	13.5***			6.46**	3.80*
R ²	.04	.16			.05	.07
▵R ²	.04	.12			.05	.02
Moderation
RSE			.19+	.06			.29***	.31***
NA × RSE				−.31**				−.11+
F			2.9*	5.5**			19.05***	2.83+
R ²			.18	.24			.15	.16
▵R ²			.02	.06			.08	.01

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was funded as part of the actions to aid the dissemination of the Research Promotion Plan of the UNED.

References

Bal

Kooij

(2011). The relations between work centrality, psychological contracts, and job attitudes: The influence of age. European Journal of Work and Organizational Psychology, 20, 497–523. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1080/13594321003669079

Bal

P. M.

De Lange

A. H.

Jansen

P. G. W.

Van der Velde

M. E. G.

(2008). Psychological contract breach and job attitudes: A meta-analysis of age as a moderator. Journal of Vocational Behavior, 72, 143–158. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1016/j.jvb.2007.10.005

Bal

P. M.

De Lange

A. H.

Jansen

P. G. W.

Van der Velde

M. E. G.

(2013). A longitudinal study of age-related differences in reactions to psychological contract breach. Applied Psychology: An International Review, 62, 157–181. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1111/j.1464-0597.2012.00524.x

Bal

P. M.

Jansen

P. G.

Van Der Velde

M. E.

de Lange

A. H.

Rousseau

D. M.

(2010). The role of future time perspective in psychological contracts: A study among older workers. Journal of Vocational Behavior, 76, 474–486. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1016/j.jvb.2010.01.002

Bamberger

P. A.

Bacharach

S. B.

(2014). Predicting retirement upon eligibility: An embeddedness perspective. Human Resource Management, 53, 1–22. doi:10.1002/hrm.21548

Coyle-Shapiro

J. A.

Conway

(2005). Exchange relationships: Examining psychological contracts and perceived organizational support. Journal of Applied Psychology, 90, 774. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1037/0021-9010.90.4.774

Fernández

M. J.

Crego

D. A.

Alcover

(2011). La transición hacia el retiro: adaptación en una muestra de prejubilados españoles de la escala Retirement Satisfaction Inventory [The transition towards retirement: Adaptation of the Retirement Satisfaction Inventory scale]. Revista Española de Geriatría y Gerontología, 46, 139–146. doi:10.1016/j.regg.2010.11.009.

Fisher

C. D.

(1997, 6). Emotions at work: What do people feel and how should we measure it? Second Biennial Australian Industrial and Organizational Psychology Conference, Melbourne, Australia. Retrieved from http://epublications.bond.edu.au/cgi/viewcontent.cgi?article=1000&context=business_pubs

Fisher

C. D.

(2000). Mood and emotions while working: Missing pieces of job satisfaction? Journal of Organizational Behavior, 21, 185–202. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1002/(SICI)1099-1379(200003)21:2<185::AID-JOB34>3.0.CO;2-M

10.

Floyd

Haynes

Rogers

Winemiller

Lemsky

Murphy

Heilman

(1992). Assessing retirement satisfaction and perceptions of retirement experiences. Psychology and Aging, 7, 609–621. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1037/0882-7974.7.4.609

11.

Fornell

Larcker

D. F.

(1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 5, 39–50.

12.

Grandey

A. A.

Tam

A. P.

Brauburger

A. L.

(2002). Affective states and traits in the workplace: Diary and survey data from young workers. Motivation and Emotion, 26, 31–55. doi:10.1023/A:1015142124306

13.

Harper

M. C.

(2005). Retirement modeling: An exploration of the effects of retirement role model characteristics on retirement self-efficacy and life satisfaction in midlife workers. Dissertation Abstracts International: Section A: Humanities and Social Sciences, 66(8), 2842.

14.

Jimmieson

(2000). Employee reactions to behavioural control under conditions of stress: The moderating role of self-efficacy. Work & Stress, 14, 262–280. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1080/02678370010015343

15.

Judge

T. A.

Ilies

(2004). Affect and job satisfaction: A study of their relationship at work and at home. Journal of Applied Psychology, 89, 661–673. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1037/0021-9010.89.4.661

16.

Newman

Jeon

Hulin

(2013). Retirement attitudes: Considering etiology, measurement, attitude-behavior relationships, and attitudinal ambivalence. In Wang

(Ed.), The Oxford handbook of retirement (pp. 228–248). New York, NY: Oxford University Press.

17.

Podsakoff

P. M.

Mackenzie

S. B.

Lee

Podsakoff

N. P.

(2003). Common method biases in behavioural research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1037/0021-9010.88.5.879

18.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.3758/BRM.40.3.879

19.

Rauschenbach

Göritz

Hertel

(2012). Age stereotypes about emotional resilience at work. Educational Gerontology, 38, 511–519. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1080/03601277.2011.567187

20.

Robinson

Morrison

E. W.

(2000). The development of psychological contract breach and violation: A longitudinal study. Journal of Organizational Behavior, 21, 525–546. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1002/1099-1379(200008)21:5<525::AID-JOB40>3.0.CO;2-T

21.

Schaubroeck

Merritt

(1997). Divergent effects of job control on coping with work stressors: The key role of self-efficacy. Academy of Management Journal, 40, 738–754. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.2307/257061

22.

Topa

Alcover

C. M.

(2015). Psychosocial factors in retirement intentions and adjustment: A multi-sample study. Career Development International, 20, 384–408. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1108/CDI-09-2014-0129

23.

Van Kleef

G. A.

(2009). How emotions regulate social life: The emotions as social information (EASI) model. Current Directions in Psychological Science, 18, 184–188. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1111/j.1467-8721.2009.01633.x

24.

Wang

Hsieh

(2014). Employees’ reactions to psychological contract breach: A moderated mediation analysis. Journal of Vocational Behavior, 85, 57–66. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1016/j.jvb.2014.04.003

25.

Weiss

H. M.

Cropanzano

(1996). Affective events theory: A theoretical discussion of the structure, causes and consequences of affective experiences at work. Research in Organizational Behavior, 18, 1–74.

26.

Zhao

Wayne

S. J.

Glibkowski

B. C.

Bravo

(2007). The impact of psychological contract breach on work-related outcomes: A meta-analysis. Personnel Psychology, 60, 647–680. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1111/j.1744-6570.2007.00087.x

Predictors	Criterion Variable: ARS		Criterion Variable: RS
	Study 1 (Hypothesis 1a)		Study 2 (Hypothesis 1b)
	β		β
	Step 1	Step 2	Step 1	Step 2
Controls
Age	−.16	−.14	.36***	.35***
Job seniority	−.02	−.00	−.29**	−.28**
Independent variable
CB		−.29***		−.19**
F	2.2	13.5***	6.5**	8.1***
R ²	.03	.11	.06	.09
▵R ²	.03	.08	.06	.03
	Criterion Variable: NA		Criterion Variable: NA
	Study 1 (Hypothesis 2a)		Study 2 (Hypothesis 2b)
	β		β
Predictors	Step 1	Step 2	Step 1	Step 1
Controls
Age	.16	.11	−.05	−.04
Job seniority	.06	.00	−.03	−.04
Independent variable
CB		.46***		.32***
F	1.43	21***	.60	23.65***
R ²	.04	.24	.006	.11
▵R ²	.04	.20	.006	.10