Month: July 2023

We again used multilevel modeling to investigate these associations. However, we were only interested in husbands’ distress as our outcome, as follows:

As indicated previously, we specified the variance-covariance matrix as autoregressive. Each distress measure (anxiety, anger, and depression) was modeled separately, with all Level 1 coefficients specified as random.

As hypothesized, wives’ anxiety was associated with greater husbands’ distress on the same day, Anxiety: b₃ = 0.11, t(32) = 3.04, p < .01; Anger/Hostility: b₃ = 0.14, t(32) = 2.84, p < .01; Depression: b₃ = 0.15, t(32) = 3.95, p < .01. There were no significant variations around the effects of wives’ anxiety on husbands’ distress (Anxiety: τ = 0.01, LR test = 0.60; Anger/Hostility: τ = 0.02, LR test = 0.20; Depression: τ = 0.01, LR test = 0.30). The LR test here represents the difference between the −2 log likelihood of a model that treats the effect of a particular coping strategy as random and a model that does not.

Read More

We identified all days on which wives reported at least moderate anxiety (2 or 3 on a 0–3 scale). We then examined the frequency, across all these days, of moderate to high (2 or 3 on a 0–3 scale) ratings on follow-up questions concerning the husbands’ influence on the wives’ anxiety that day. Wives reported at least moderate anxiety on an average of 10 of 14 days (range = 6–14). On 60.5% of these occasions, wives perceived their husbands as having at least some (ratings of 2 or 3 on a 0–3 scale) influence on their anxiety that day. On 44.4% of these days, husbands were perceived as improving the wife’s anxiety; on 17.4% of days, husbands were perceived as making the wife’s anxiety worse; and on 38.4% of days, husbands were perceived as neither making the anxiety better nor worse. A one-way chi-square test revealed significant differences in the proportions of each type of rating (i.e., anxiety made better, worse or neither) represented across all high-anxiety days, χ²(2, N = 66) = 36.69, p < .05.

Read More

We tested the moderation hypothesis in two separate multilevel analyses. The same model described in Equation 1 was used for each analysis. However, Level 2 equations included the following additional predictors as moderators: CIQ and DAS. Because we were interested in the effects of wives’ anxiety on wives’ and husbands’ relationship quality (denoted as b_3w and b_3h in Equation 1, respectively), b_3w, and b_3h were modeled as a function of these moderators, as such:

CIQ_wi reflects wives’ report of anxiety-specific relationship adjustment for dyad i, and DAS_wi reflects wives’ report of global relationship functioning. We did not alter the specifications of the random effects for the moderation tests.

CIQ scores moderated the within-person association between wives’ anxiety and their own ratings of positive relationship quality that same day, γ_31w = 0.017, t(32) = 2.57, p < .05, such that for wives who reported low CIQ scores, their anxiety was negatively associated with their positive RQ. For wives who reported high CIQ scores, their anxiety was not associated with their positive RQ. Figure 1 is the visual representation of this interaction, for CIQ scores one standard deviation above and below the mean. CIQ scores did not moderate any other hypothesized within-dyad associations, Positive RQ: γ_31h = 0.001, t(32) = 0.17, ns; Negative RQ: γ_31w = −0.006, t(32) = −0.89, ns; γ_31h = 0.005, t(32) = 0.79, ns. As hypothesized, DAS did not moderate the association between anxiety and negative or positive RQ, Positive RQ: γ_32w = 0.005, t(32) = 0.60, ns; γ_32h = 0.004, t(32) = 0.52, ns; Negative RQ: γ_32w = −0.011, t(32) = −1.20, ns; γ_32h = −0.010, t(32) = −1.24, ns.

Read More

We used a multilevel statistical model to investigate these associations separately for positive and negative RQ. Multilevel modeling accounts for the bias in standard errors and statistical tests that result from nonindependent data and effectively handles “unbalanced” or missing data at the level of repeated observations by using all available data for participants. Inferences are valid assuming missing data are missing at random (see Schafer & Graham, 2002). The models had two levels: within-dyad (over time) and between-dyad. Using the dyadic longitudinal approach described by Bolger and Shrout (2007), we included wives’ and husbands’ RQ in a single multilevel analysis to account for the fact that wives’ and husbands’ data were clustered within dyad. All analyses were conducted using the MIXED procedure in SAS (Version 9.1.3, 1997).

The within-dyad level of the analysis allowed each dyad’s RQ to be modeled as a function of husbands’ and wives’ anxiety. We predicted a given day’s husbands’ and wives’ RQ for a particular dyad and adjusted for number of days in the study and weekend effects. Because husbands’ anxiety may be highly associated with husbands’ evaluations of RQ, husbands’ anxiety was included to adjust for this effect. The model specified was as follows:

where Y_ijk is the RQ for dyad i for person j (j = 1 is wife’s report; j = 2 is husband’s report) on day k. When the outcome is the wife’s report (Wife_ijk = 1 and Husb_ijk = 0), the first part of the model is selected and all of the b coefficients have the subscript w. Similarly, when the outcome is the husband’s report, Wife_ijk = 0, Husb_ijk = 1, and the second part of the model is selected. Days_ik is the number of days in the study; Weekend_ik indicates whether it is a weekend day or not; WAnx_ik is the wife’s report of anxiety; HAnx_ik is the husband’s report of anxiety; the residual components are represented by e_ijk. All predictor variables were within-person centered (Raudenbush & Bryk, 2002). Finally, the approach discussed by Bolger and Shrout (2007) allowed us not only to account for dependency within individuals across time (i.e., autoregressive) but also to account for dependency within dyads (pp. 292–295).

The between-dyad level of this analysis modeled individual differences in the coefficients specified in Equation 1. We fit a model that considered intercepts for both wives’ and husbands’ reports of RQ to be random (i.e., varying across persons). In addition, slope of day on wives’ RQ (b_1w), slope of wives’ anxiety on wives’ RQ (b_3w), and slope of husbands’ anxiety on husbands’ RQ (b_4h) were modeled to be random for the positive RQ analysis; slope of husbands’ anxiety on husbands’ RQ (6_4h) was modeled to be random for the negative RQ analysis.¹ Random effects were tested using the nested comparison of likelihood ratio (Singer & Willett, 2003, p. 119).

Table 3 presents results for both wives’ and husbands’ reports of positive RQ.² Only variables of interest are reported here. Wives’ anxiety was not associated with their own positive RQ, b_3w = −0.03, t(32) = −0.89, ns, but was significantly associated with husbands’ positive RQ, b_3h = −0.14, t(32) = −3.05, p < .01. Specifically, on days when wives experienced higher anxiety, husbands reported less positive relationship quality. There was no significant variation around these effects.

Read More