Capturing life as it is truly lived? Improving diary data in educational research

Figures & data

Table 1. Logistic regression predicting likelihood of participating in the ISLE diary.

	B	SEB	Wald	df	p	Odds ratio	95% CI for Odds ratio
							Lower	Upper
Gender	.74	.28	6.86	1	.009	2.10	1.21	3.66
English proficiency (C-tests)^a	.70	.25	8.07	1	.004	2.01	1.24	3.25
Attitudes towards ISLL^a	−.02	.09	.08	1	.789	.98	.82	1.16

^a These analyses are based on standardized factor scores. Thus, the odds ratios indicate the changes in likelihood of participating in the diary portion of the study per SD increase in the C-test/attitudes scores.

The model had statistically significant predictive power (χ²[3] = 18.40, p<.001), explained 8.6% of the variance in the dependent variable (Nagelkerke R²), and correctly classified 76% of the cases.

Table 2. Multiple regression predicting number of days participating in the ISLE diary.

	B	95% CI for B		SEB	β	t	p
		Lower	Upper
Gender	.88	.33	1.42	.28	.20	3.15	.002
Prior proficiency (C-Test factor score)	.52	.03	1.01	.25	.14	2.09	.038
Attitudes towards ISLL (factor score)	−.19	−.36	−.02	.09	−.15	−2.22	.027

The model had significant predictive power (F[3,235] = 6.88, p<.001) and explained 6.9% of the variance in the dependent variables (adjusted R²).