A process-oriented perspective examining the relationships among daily coping, stress, and affect

Abstract

We assess the daily relationships between age, three functional coping strategies, and positive and negative affect while accounting for the individual’s cognitive appraisal of the severity and controllability of their encountered stressor. We collected 56 days of daily data from participants in the Notre Dame Study of Healthy & Well-Being assessing affect, the most bothersome event experienced each day, and coping strategies used to deal with that event (N = 371; Age 44–87; M = 67.41; SD = 8.87). Multi-level modeling allowed us to explore and compare the between- and within-person effects. The main effects revealed that coping strategies relate to affect differently. The interaction terms revealed that the effectiveness of each coping strategy depended on characteristics of the encountered stressor and/or characteristics of the individual. Average values of stress severity, Altering the Meaning, and Dispelling the Effects more strongly related to PA or NA than daily fluctuations on these constructs. Findings illustrate that certain coping strategies target affect differently. Approaching research questions regarding stress and coping with a process-oriented perspective, through the use of daily data, allows for a more thorough understanding of the real-time, lived relationships among the individual, stress, and coping.

Evolutionarily, stress motivates and protects individuals from hazardous environmental conditions, but as individuals experience more stress in their everyday lives, this protective resource extracts a toll on well-being (Sapolsky, 2004). That is, without proper modulation, stress can lead to an increased risk of negative emotional and physiological consequences due to its detrimental effects on neuroendocrine, cardiovascular, immune, and metabolic systems. In fact, research illustrates that greater affective reactivity, or the daily relationship between stress and negative affect (NA), predicts emotional (Charles, Piazza, Mogle, Sliwinski, & Almeida, 2013) and physical health (Piazza, Charles, Sliwinski, Mogle, & Almeida, 2012) as much as 10 years later. Effective Coping protects individuals from the negative effects of stress by disrupting the link between stress and well-being (Pearlin & Schooler, 1978); however, the effectiveness and importance of certain coping strategies may change as individuals age (Aldwin, Skinner, Zimmer-Gembeck, & Taylor, 2011). Because individuals differ in the way they experience the stress-affect relationship and use coping strategies, it is important to examine these relationships at the daily level in order to account for intraindividual variability and individual differences in that variability (Stone & Neale, 1984). The current study uses 56 days of daily data to examine the dynamic relationships between age, stress, and coping on NA and positive affect (PA), which can ultimately contribute to the development of interventions and preventative care aimed toward disrupting the link between daily affective reactivity and negative psychological and physiological outcomes.

Stressful encounters begin with a two-part cognitive appraisal, during which the individual first appraises how much their situation threatens their well-being, termed primary appraisal, and then determines what, if anything, can be done to modulate their stressful situation, termed secondary appraisal (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986). This appraisal process indicates that the severity of the stressor at hand (assessed during primary appraisal), as well as the controllability of the stressor (assessed during secondary appraisal) influence how individuals perceive their stressful circumstances and ultimately how they choose to cope with them. Researchers emphasize the need to study coping from a process-oriented perspective by gathering contextual information about the stressor, the individual, and the coping strategy (Folkman et al., 1986; Lazarus & Folkman, 1984; Tennen, Affleck, Armeli, & Carney, 2000), indicating that certain coping strategies can be effective in some situations but not others. For example, the improvements in emotional regulation that older adults experience to compensate for diminished physical or cognitive functioning (Urry & Gross, 2010) may not only influence the way they cope with stress over time (Aldwin et al., 2011), but reflect age differences in coping effectiveness. Consequently, coping cannot be separated from the context in which it arises, meaning both the person and environmental situation impact the coping process (Folkman et al., 1986). The current study will explore the contextual factors of age, perceived stress severity, and perceived stress controllability and their relationships with various coping strategies and affect.

Although there are many different types of coping strategies, Pearlin and Schooler (1978) argue that coping has three functional purposes. First, individuals might Alter the situation (Resituating), which involves focusing on the specific problem at hand and working to dispel that problem. Second, individuals might Alter the meaning (Meaning-making) by cognitively reducing the impact of the encountered stressor. Third, individuals might Dispel the effects (Dispelling) of stress by working to dissipate the emotional and physiological consequences associated with stress, which can help the individual adapt to their stressful circumstances without necessarily changing them or changing the meaning of them. Theorists suggest that certain functions of coping might be more effective in dealing with a specific stressor than others (Aldwin et al., 2011; Folkman et al., 1986; Pearlin & Schooler, 1978). For example, engaging in Resituating in response to an uncontrollable stressor may be less effective than Meaning-making in response to that stress. Empirically testing the contextual nature of these relationships requires information regarding both the encountered stressor and the coping strategies used, which can be tested with an idiographic perspective by examining these relationships in prospective time.

Researchers emphasize the need to study coping from a process-oriented perspective by gathering contextual information about the stressor, the individual, and the coping strategy. (Folkman et al., 1986; Lazarus & Folkman, 1984; Tennen et al., 2000). By using an idiographic approach, through which relations between variables are observed within individuals over time, researchers can better test theories of coping that illustrate coping as a complex and dynamic process, dependent on individual characteristics and the stressor at hand (Molenaar & Campbell, 2009; Tennen et al., 2000). Research illustrates differences in retrospective and prospective strategies designed to capture the construct of coping (Ptacek, Smith, Espe, & Raffety, 1994; Tennen et al., 2000; Todd, Tennen, Carney, Armeli, & Affleck, 2004). Specifically, individuals have a tendency to over-report the effectiveness of coping in retrospective reports (Ptacek et al., 1994). Moreover, dispositional coping does not effectively predict intraindividual variability in coping or interindividual differences in that variability (Todd et al., 2004). These results suggest that the construct of coping may be too complex to study at the global level, and coping cannot be understood when removed from its related stressor.

Although much research has established the relationship between daily stress and affect (Charles et al., 2013; Sliwinski, Almeida, Smyth, & Stawski, 2009; Stawski, Sliwinski, Almeida, & Smyth, 2008; Piazza et al., 2012), less research has examined how coping impacts these relationships (Tennen et al., 2000). Daily designs are a particularly useful tool that can help researchers understand coping in real time, over multiple occasions, and in relation to specific stressors (Stone, Kennedy-Moore, & Neale, 1995; Tennen et al., 2000). Diary methods (Bolger, DeLongis, Kessler, & Wethington, 1989) capture processes as they unfold, reflecting the “lived” experience of everyday life (Wheeler & Reis, 1991). A fundamental benefit of this method is that it allows researchers to assess reported events and experiences in their natural, spontaneous context and provides complementary information to that acquired through more traditional designs (Bolger et al., 1989). A second benefit is the reduction in retrospective bias that may occur if the amount of time that elapses between an experience and the account of this experience is long. In this way, researchers can better understand the ebb and flow of the relationship between stress and coping.

Because between-person findings cannot be generalized to the within-person level (Molenaar & Campbell, 2009), the current study is largely exploratory in order to understand the relationships between age, stress, coping, and affect. The first aim of the current study is to understand the relationship between Pearlin and Schooler’s (1978) three functional aspects of coping and PA and NA. The second aim is to examine the extent to which the interaction between age, primary appraisal, and secondary appraisal with specific coping strategies relates to changes in affect. The final aim is to compare the impact of the between- and within-person effects of cognitive appraisal and coping on PA and NA.

1. Method

1.1. Sample

There were 371 participants from Wave 7 of the Notre Dame Study of Health & Well-Being (NDHWB), ranging in age from 44 to 87 (M = 67.41, SD = 8.87) recruited from a five-county region in Indiana. The NDHWB is a longitudinal study that conducts global and 56-day daily burst assessments of individual physical and emotional health (see Bergeman & Deboeck, 2014). The sample was 64% female and predominantly White (84%). The sample was representative of the region from which it came, with a diverse range of education (34% reported high school or lower as their highest level of education, and 34% reported having a college degree or higher) and income (20% reported making less than $15,000, and 41% reported making more than $40,000).

1.2. Procedure

After completing the Year 7 global questionnaire from the NDHWB, participants were invited to participate in the 56-day daily “burst” assessment, which was the wave that included a measure of daily coping strategies in relation to daily stress. Participants were instructed to complete the questionnaire at the end of each day. Participants received a total of $80 in $10 increments in exchange for mailing back each week of daily questionnaires. Because there were 371 participants and 56 days in the diary burst, 20,776 days of data were possible. Participants only rated their coping strategies on days that they endorsed a bothersome event; this occurred on 11,202 (54%) of the days.

1.3. Measures

1.3.1. Daily coping instrument (DCI)

The perceived severity of the encountered stressor, perceived controllability of the encountered stressor, and functional coping strategies were assessed with the DCI (Stone & Neale, 1984). The participants ranked the most bothersome event they experienced that day in terms of severity and controllability on a scale of 1–10 with 10 being most stressful or most controllable (see Supplemental Table 1 for the types and percentages of stressors endorsed). They marked which coping categories they utilized to deal with the bothersome event. The coping categories differed slightly from Stone and Neale’s original measure in that they included examples and modified wording to increase clarity for use in an older population. A factor analysis (see Supplemental Table 2 for details) revealed that three factors were present in the data, illustrative of Pearlin and Schooler’s (1978) coping functions of 1) Resituating, 2) Meaning-making of stress, and 3) dispelling the negative effects of stress. Resituating included one item scored 0 if not endorsed or 1 if endorsed (“I accepted the issue for what it was and moved on with my day”). Meaning-making included three items scored 0 to 3 depending on how many of the items the participant endorsed: talking to a trusted person, gathering information about the issue, or reframing the issue. Dispelling included three items similarly scored 0 to 3: distracting oneself, engaging in physical activity, or doing something relaxing. A Durbin-Watson test revealed that the lag-1 autocorrelation was 0.05 for Resituating (d = 1.90), 0.10 for Meaning-making (d = 1.79), and 0.09 for Dispelling (d = 1.82).

1.4. Positive and negative affect schedule (PANAS)

Daily responses on the PANAS (Watson, Clark, & Tellegen, 1988) were used to assess PA and NA. Participants reported the extent to which they felt each of 10 emotions on that day using a scale from 1 to 5 (not at all, a little, moderately, quite a bit, extremely). The measure assessed emotions such as “ashamed,” and “guilty,” “excited,” and “interested.” Cronbach’s alpha on Day 1 was 0.90 and 0.94 for the NA and PA subscales respectively. A Durbin-Watson test revealed that the lag-1 autocorrelation was 0.76 for NA (d = 0.48) and 0.81 for PA (d = 0.39).

2. Analytic approach

We used person-mean centering to remove the between from the within-person effects from the time-varying covariates (Curran & Bauer, 2011; Wang & Maxwell, 2015) and included time as a predictor variable at Level 1 to control the linear trend in the independent variables (Wang & Maxwell, 2015). We contrasted the within-person main effects to the between-person main effects to explore whether the associations with NA are stronger at the intra- or inter-individual level (Wang & Maxwell, 2015). The data were analyzed using the SAS Proc Mixed procedure with Maximum Likelihood, so missing data was assumed to be missing at random or missing completely at random.

We assessed the effect of each aspect of cognitive appraisal separately to understand the unique relationship between these aspects of cognitive appraisal and affect. The first set of analyses explored the relationships among daily coping, daily stress severity, and NA and PA. At Level 1, each predictor varies according to the individual’s (i) day (j). The models were built sequentially, first assessing the main effects, then the two-way interactions, and then the three-way interactions. The main effects model controls for time (γ₁₀) and assesses both the within-person (γ₂₀, γ₃₀, γ₄₀, γ₅₀) and between-person main effects (γ₀₁, γ₀₂, γ₀₃, γ₀₄, γ₀₅) of the severity of the encountered stressor, and each coping strategy on NA and PA. We then added the level-1 by level-2 interactions between age and severity and age and each strategy (γ₂₁, γ₃₁, γ₄₁, γ₅₁) as well as the within-person interactions between stress severity and each coping strategy (γ₆₀, γ₇₀, γ₈₀). Finally, we added the level-1 by level-2 three-way interactions between age, severity, and each coping strategy (γ₆₁, γ₇₁, γ₈₁) to explore the relationships between age, stress severity, coping, and affect. The interactions are all specified as fixed in order to make the model positive definite. The second set of analyses replaced stress severity with stress controllability.

3. Results

3.1. Descriptive statistics

There were no gender differences on NA, PA, or stress controllability. The sample did illustrate gender differences on stress severity, with women reporting greater stress severity (M = 6.57, SD = 1.89) than men (M = 5.70, SD = 1.82; t = −4.22, p < 0.001; Cohen’s d = 0.47). In addition, men reported greater use of Resituating (M = 0.74, SD = 0.22) compared to women (M = 0.67, SD = 0.25; t = 2.58, p = 0.010; Cohen’s d = 0.30).

3.2. Stress severity

The results of the main effects model (Table 1) revealed that age positively related to daily PA. Daily stress severity and overall between-person (mean effect) stress severity related to daily NA, but the between-person effect was a stronger predictor (γ₀₂–γ₂₀ = 0.28, p = 0.006). Both daily stress severity and between-person stress severity were also associated with less daily PA. The between-person (mean) effect of stress severity was a stronger predictor of PA (γ₀₂–γ₂₀ = −0.54, p = 0.009). Daily Resituating was associated with less daily NA and greater daily PA. Higher overall levels of Resituating were associated with lower NA. Daily Meaning-making was associated with greater daily NA and greater PA. Between-person (mean) Meaning-making was associated with greater daily PA and was a stronger predictor than the within-person effect PA (γ₀₄–γ₄₀ = 2.57, p < 0.001). Daily Dispelling was associated with greater NA as was the between-person (mean) effect, which was a stronger predictor (γ₀₅–γ₅₀ = 1.47, p < 0.001).

Table 1.

The main effects of age, stress severity, and each coping strategy on affect. A Pseudo R² revealed that, compared to an intercept only model, the main effects model explained 16% of between-person variance and 13% of within-person variance in daily PA. The main effects model explained 14% of the between-person variance in daily NA, but the within-person variance was not interpretable due to a negative Pseudo R² (Singer & Willett, 2003).

	DV: PA			DV: NA
	Fixed effects	SE	t value	Fixed effects	SE	t value
Fixed within effects estimates
γ00 (intercept)	25.69***	3.38	7.59	10.81***	1.62	6.66
γ10 (time)	−0.02***	0.01	−3.56	−0.01*	0.00	−2.29
γ20 (stress severity)	−0.22***	0.03	−7.30	0.30***	0.02	14.31
γ30 (situation)	0.41**	0.14	3.03	−0.50***	0.04	−5.39
γ40 (meaning)	0.44***	0.07	6.02	0.17***	0.04	4.43
γ50 (dispel)	0.03	0.09	0.36	0.10*	0.05	2.00
Fixed between effect estimates
γ01 (age)	0.13**	0.04	3.10	−0.02	0.02	−1.09
γ02 (stress severity)	−0.76***	0.21	−3.70	0.58***	0.10	5.83
γ03 (situation)	−0.61	1.63	−0.38	−1.66*	0.79	−2.10
γ04 (meaning)	3.01***	0.70	4.28	0.19	0.34	0.56
γ05 (dispel)	0.04	0.74	0.05	1.58***	0.35	4.48

^*p < 0.05.

^**p < 0.01.

^***p < 0.001.

Adding the two-way interactions revealed that greater age buffered the inverse relationship between stress and PA (γ₂₁ = 0.01, p < 0.001). In terms of NA, during times of low stress, greater Resituating related to greater NA, but during times of high stress, situational coping related to lower NA (γ₆₀ = −0.12, p < 0.001). In addition, greater Meaning-making related to an exacerbated relationship between stress severity and NA (γ₇₀ = 0.06, p < 0.001). The final model revealed a significant three-way interaction between age, stress severity, and Meaning-making (γ₇₁ = −0.01, p = 0.002), with age buffering the positive relationship between greater meaning-focused coping and NA when individuals experienced greater daily stress. During times of low stress, greater age buffered the inverse relationship between low meaning-focused coping and NA.

3.3. Stress controllability

The results of the main effects models after replacing stress severity with stress controllability showed that control was associated with lower NA within individuals (γ₂₀ = −0.10, p < 0.001). Greater daily stress controllability was associated with higher daily PA (γ₂₀ = 0.12, p < 0.001) as were greater overall levels of stress controllability (γ₀₂ = 0.75, p < 0.001), and the between-person effect was a stronger predictor of PA than the within person effect (γ₀₂–γ₂₀ = 0.63, p = 0.003). The main effects of each coping strategy show a similar pattern as the main effects with stress severity, although the parameter estimates differ slightly (Table 2). When the two-way interactions were included in the model, age buffered the impact of low daily controllability on NA (γ₂₁ = 0.003, p = 0.015). In addition, daily Resituating interacted with stress controllability to predict NA (γ₆₀ = 0.10, p < 0.001), which revealed that daily use of high situational coping relates to lower NA, and high daily controllability buffers the effect of low situational coping on NA. The results of the three-way interactions revealed that during times of low controllability, greater dispelling relates to less NA and higher age mitigates the effect of low dispelling on NA. During times of high controllability, however, greater dispelling leads to greater NA and higher age buffers this effect (γ₈₁ = −0.004, p = 0.043).

Table 2.

The main effects of age, stress controllability, and each coping strategy on affect. A Pseudo R² revealed that, compared to an intercept only model, the main effects model explained 13% of the between-person variance and 12% of the within-person variance in daily PA. The main effects model explained 4% of the between-person variance in daily NA, but the within-person variance was not interpretable due to a negative Pseudo R² (Singer & Willett, 2003).

	DV: PA			DV: NA
	Fixed effects	SE	t value	Fixed effects	SE	t value
Fixed within effects estimates
γ00 (intercept)	16.73***	3.03	5.52	16.30***	1.51	10.78
γ10 (time)	−0.02***	0.01	−3.38	−0.01	0.00	−1.92
γ20 (stress controllability)	0.12***	0.02	6.02	−0.10***	0.01	−7.93
γ30 (situation)	0.60***	0.14	4.19	−0.72***	0.10	−7.39
γ40 (meaning)	0.35***	0.08	4.59	0.25***	0.04	5.74
γ50 (dispel)	0.02	0.09	0.27	0.16***	0.06	2.79
Fixed between effect estimates
γ01 (age)	0.13**	0.04	3.19	−0.03	0.02	−1.50
γ02 (stress controllability)	0.75***	0.21	3.63	−0.13	0.10	−1.26
γ03 (situation)	1.07	1.55	0.69	−3.21***	0.78	−4.11
γ04 (meaning)	0.35***	0.08	4.59	0.35	0.36	0.98
γ05 (dispel)	0.02	0.09	0.79	1.90***	0.36	5.20

^**p < 0.01.

^***p < 0.001.

4. Discussion

The results provide a unique lens for illustrating how daily increases or decreases in stress and coping for any one individual relate to PA and NA and how these relationships differ among those of different ages. Specifically, the within-person findings illustrate how individuals compare to themselves by indicating effects when participants are above or below their own personal average of perceived intensity of stress severity, amount of stress controllability, or frequency of use of each coping strategy. The results corroborate previous literature by revealing that within-person fluctuations in stress severity negatively relate to PA (Stawski et al., 2008) and positively relate to NA (Charles et al., 2013; Sliwinski et al., 2009; Stawski et al., 2008; Piazza et al., 2012) whereas within-person daily increases in stress controllability positively relate to PA and negatively relate to NA. The within-person effects of the three functional coping strategies on affect uniquely add to the literature in that they overcome the retrospective bias associated with more global designs (Ptacek et al., 1994; Tennen et al., 2000; Todd et al., 2004), and reflect the lived experience of encountering and coping with stress (Wheeler & Reis, 1991).

The within-person main effects of the three functional coping strategies on affect revealed that different coping functions relate to affect differently (Aim 1). Specifically, when individuals used Resituating, they experienced lower daily NA and higher daily PA whereas using more Meaning-making strategies than usual related to both higher daily NA and higher daily PA. It is possible that Resituating either reduces NA over the course of one day, or is typically used on days when the individual experiences less NA. On the other hand, it is possible that individuals use Meaning-making more when they feel more emotionally complex. For example, when individuals use this strategy, they may be able to see the good in the bad as they experience distress from their stressful circumstances. This interpretation is consistent with theoretical work by Folkman and Moskowitz (2000), suggesting that individuals can experience PA during times of stress, and this PA ultimately benefits the individual. Although Meaning-making might not necessarily directly reduce NA, research shows that PA promotes better emotional and physiological health, essentially “undoing” the negative ramifications associated with NA (Fredrickson, 2001). Consequently, Meaning-making might promote greater PA, which benefits the individual in the long-term.

The within-person interactive effects demonstrate that the affective impact of engaging in each coping strategy depends on aspects of the encountered stressor as well as aspects of the individual (Aim 2), highlighting the importance of taking into account the context in which coping arises (Folkman et al., 1986; Lazarus & Folkman, 1984; Tennen et al., 2000). Specifically, the effect of Resituating on NA changed when accounting for stress severity and controllability. In addition, accounting for individuals’ ages altered the interaction between Meaning-making and stress severity as well as Dispelling and stress controllability on NA. Accounting for the way in which older adults cope with stress may explain the greater intraindividual variability in the stress-NA relationship found in previous literature (Sliwinski et al., 2009). These findings suggest that older adults may experience the relationships between stress, coping, and affect differently than middle-aged adults, due to possible differences in stress (Pearlin & Skaff, 1996) and emotional regulation (Urry & Gross, 2010).

MLM also allows us to compare the relative impact of the between-person effects to the within-person effects (Aim 3; Wang & Maxwell, 2015). In this way, we can assess whether average differences between individuals or daily fluctuations in stress and coping are more strongly related to PA and NA. Overall levels of Dispelling and Meaning-making more strongly positively related to NA and PA respectively than daily fluctuations, suggesting that individuals who use these coping strategies more than others experience greater daily NA or PA compared to an individual using this strategy more than they usually do. These findings suggest that coping effectiveness depends on the user, again highlighting the contextual nature of the coping process (Folkman et al., 1986). In addition, the strength of the between-person effects compared to the within-person effects suggests that coping frequency plays a role in the relationship between stress and affect. Specifically, the more individuals use these coping strategies the stronger the relationships between coping and affect. Furthermore, certain coping strategies may be better suited for certain types of individuals. Future research can examine whether certain personality characteristics impact the stress-coping-affect relationships.

Because the current study examines stress, coping, and affect within the same day, the directionality of the results remains unclear. For example, the coping strategies might buffer the impact of stress severity on NA within an individual, but this buffering effect may take place over the course of several days. Similarly, it is possible that the PA or NA an individual experiences elicits certain coping strategies. Although Meaning-making related to higher PA, it also related to higher NA. Therefore, individuals might engage in Meaning-making in response to their higher amounts of NA compared to Resituating. In addition, although there may be a positive relationship between Meaning-making and NA within one day, it may be that there is a negative relationship between these constructs when examined with lagged analyses across days within individuals. Future research can examine the bidirectionally lagged relationships between affect and coping to better understand the processes by which these constructs operate.

The 56-days of daily data and large sample size not only add strength to the study’s findings, but allow us to test the relationships among stress, coping, and affect in a more novel, idiographic way. The study, however, does have some limitations. Although the sample is representative of the region from which it came, the study’s generalizability may be limited to the 5-county region in Northern Indiana from which it was drawn; it is possible that using a sample from a different geographical location could yield different results. In addition, there are many different ways to conceptualize coping in the literature (e.g. emotion- versus problem-focused coping (Lazarus & Folkman, 1984)), so using different coping measures might reveal different findings.

The current study does uniquely combine foundational theoretical work on stress and coping by considering how the cognitive appraisal of stress (i.e. stress severity and controllability), described by Folkman et al. (1986) relates to the three functional aspects of coping described by Pearlin and Schooler (1978). Moreover, the study overcomes potential problems with between-person retrospective coping data in its use of daily data (Ptacek et al., 1994; Tennen et al., 2000; Todd et al., 2004). Specifically, we could examine the relationships among stress, coping, and affect with a process-oriented, prospective perspective, thereby highlighting the impact of person-related, day-related, and stress-related contextual factors on the coping-affect relationship. By parsing within-person variability from between-person differences, the current study demonstrates the influence of both daily fluctuations in stress and coping on affect as well as the impact of individual differences, including age and average differences in stress and coping, on affect, ultimately illustrating the contextual nature of the stress and coping process. Because the daily stress-NA relationship impacts long-term emotional and physiological health (Charles et al., 2013; Piazza et al., 2012), understanding how contextual factors impact this relationship, like coping, can better inform intervention and preventative care, and ultimately contribute to fostering better emotional regulation and overall healthier well-being.