Initial Analyses
As additional manipulation checks, two ples t tests were conducted to examine differences in ITRS scores. The results confirmed that participants assigned to the growth condition reported stronger growth beliefs (M = 5.87, SD = 0.74) than did those in the destiny condition (M = 5.52, SD = 1.01), t(302) = 3.61, p < .001, d = 0.40. Participants assigned to the destiny condition also reported stronger destiny beliefs (M = 4.75, SD = 1.12) than did those in the growth condition (M = 3.92, SD = 1.18), t(302) = 6.22, p < .001, d = 0.72.
The effect from implicit theories out of matchmaking with the cheating forgiveness
To examine whether the type of behaviour (H1), the sex of the forgiver (H2), and the manipulation of ITRs affected infidelity forgiveness (H5), a 2 (experimental condition; growth/destiny) ? 2 (sex of forgiver) ? 4 (type of behaviour) mixed-design ANOVA was conducted. A significant main effect of type of behaviour emerged, F(1.73, ) = , p < .001, ?p 2 = .75. Consistent with Study 1 (and H1), multiple comparisons indicated that all subscales were significantly different from one another (ps < .001; See Table 1). Consistent with Study 1 (partially consistent with H2), a significant main effect of sex of forgiver also emerged, F(1, 232) = , p < .001, ?p 2 = .09, in which male participants forgave to a greater extent (M = 4.41, SD = 1.15) than did female participants (M = 3.73, SD = 1.00).
As expected (H5), the results also indicated that there was a significant main effect of experimental condition, F(1, 232) = , p < .001, ?p 2 = .06; those in the growth condition forgave their partner's hypothetical infidelity to a greater extent (M = 4.33, SD = 1.12) than did those in the destiny condition (M = 3.80, SD = 1.02). Interestingly, this main effect was qualified by two significant two-way interactions. The first significant interaction occurred between condition and type of behaviour, F(1.58, ) = , p < .001, ?p 2 = .03. Simple effects analysis revealed that the effect of the experimental condition was only significant for the emotional/affectionate behaviours, F(1, 316) = , p = .002, ?p 2 = .03, and the solitary behaviours, F(1, 316) = , p = .001, ?p 2 = 0.04. When forgiving a partner's hypothetical emotional/affectionate and solitary behaviours, those receiving the growth manipulation forgave to a greater extent than those receiving the destiny manipulation (see Figure 1).
The following one or two-ways interaction happened anywhere between position and you can gender, F(1, 301) = 5.60, p = .02, ?p dos = .02. Effortless outcomes study indicated that this new manipulation is actually extreme to have men users, F(step 1, 301) = 7.22, p = .008, ?p dos = .02, however women people, F(step one, 301) = 0.05, p = .82, ?p 2 = .00. Among male participants, those who work in the development status forgave the lover’s hypothetical cheating to an elevated extent than simply did those in the fresh fate reputation (see Contour 2). This new control did not apply at lady participants’ infidelity forgiveness. Few other two- otherwise about three-way interactions abilities was basically tall. Footnote step 1
Examining dispositional attachment insecurity since the a moderator
To assess H6, four hierarchical several regression analyses hookup Raleigh reddit was indeed used where the ECRS subscale results was in fact joined towards first step, the fresh new dummy coded experimental updates into step two, and the ECRS ? standing telecommunications terminology towards step three. The newest DIQ-R subscales was integrated as lead parameters (shortly after centred to attenuate multicollinearity). Due to the fact a beneficial Bonferroni modification was applied to guard out-of form of I mistakes, a leader from .01 (.05/4) are followed. Discover Desk 3 getting correlations.
Recent Comments