Table 6.

The relationships within the intervention condition between strategy use and the IAT, concern, discrepancies, and the components of discrepancies, shoulds and woulds.

	Predictor	IAT	Concern		Discrepancies		Shoulds		Woulds
Replacement	Average	−.014 [−.181, .152]	−.303 [−1.450, .846]		.284 [−.411, .978]		−.109 [−.614, .395]		.141 [−.626, .899]
	Change	.030 [−.044, .108]	−.026 [−.272, 220]		.283 [.040, .542]	*	−.146[−.298, −.010]	*	.048 [−.138, .248]
Imaging	Average	.010 [−.166, .187]	−.944 [−2.159, .273]		.380 [−.360, 1.118]		.050 [−.485, .585]		.398 [−.410, 1.207]
	Change	−.007 [−.085, .101]	−.340 [−692, −.020]	*	.121 [−.068, .345]		.017 [−.127, .193]		.168 [−.050, .405]
Individuating	Average	−.044 [−.135,.047]	−.501 [−1.123,.122]		−.048 [−.430, .334]		−.145 [−.421, .131]		−.181 [−.596, .234]
	Change	.024 [−.018, .070]	.079 [−.100, .268]		−.034 [−.141, .068]		−.015 [−.098, .057]		−.049 [−.180, .038]
Persp. taking	Average	−.081 [−.238, .077]	.279 [−.709, 1.469]		−.131 [−.788, .536]		−.057 [−.531, .416]		−.216 [−.936, .502]
	Change	.052 [−.014, .135]	.146 [−.169, 499]		.016 [−.189, .231]		−.084 [−.209, .032]		−.104 [−.243, .035]
Contact	Average	−.139 [−.294, .017]	.267 [−.822, 1.357]		−.117 [−.774, .539]		−.089 [−.562, .384]		−.190 [−.909, .528]
	Change	.021 [−.072, .121]	−.094 [−.376, .171]		.068 [−.100, .241]		−.033 [−.155, .087]		.035 [−141, .229]
All strategies	Average	−.019 [−.055, .017]	−.100 [−.352, .152]		.009 [−.143, .162]		−.035 [−.145, .075]		−.028 [−.194, .139]
	Change	.019 [−.008, .049]	−.004 [−.089, .086]		.024 [−.029, .092]		−.021 [−.056, .012]		−.010 [−.051, .036]

Note: Each strategy use variable was derived by dividing the number of uses of the strategy by the time since the last measurement. We then used both each participant’s average daily rate and the differences between their occasion-specific daily rates and this average as predictors in an LMEM, along with a random slope for the centered occasion-specific rates. All 95% CIs were derived using profile likelihood. All models also contained effects of condition, linear and quadratic time, a random intercept for each participant, and random slopes for linear and quadratic time.