From quality to outcomes: a national study of afterschool STEM programming

International Journal of STEM Education

Table 4 Mean (± SD) change in retrospective pretest-posttest ratings for nine CIS-S scales with between-state comparisons

	Retro pretest		Retro posttest		Stat.		Signif.	Effect size	Proportion of changes			Between-groups	Effect size
Variable	M	SD	M	SD	t	df	p	d	Pos. (+)	Neut. (=)	Neg. (−)(−)	States	\( {\eta}_P^2 \)
STEM engagement	60.2	19.9	70.0	19.3	26.6	1598	< 0.001	0.50	77.5%	3.5%	18.9%	1:11*	.017
STEM career interest	51.0	22.0	59.9	21.9	25.3	1598	< 0.001	0.40	75.7%	3.7%	20.6%	1:11*	.017
STEM career knowledge	45.0	23.3	55.7	21.5	28.6	1598	< 0.001	0.49	79.7%	3.1%	17.1%	1:3, 1:5; 1:7; 1:11; 2:7; 3:9; 5:9; 7:9; 11*:9	.028
STEM activity participation	36.5	21.2	45.3	22.4	26.8	1598	< 0.001	0.40	76.7%	3.5%	19.8%	1:10; 1:11; 6:10*	.018
STEM identity	50.2	22.8	58.1	22.3	23.9	1598	< 0.001	0.35	73.1%	4.0%	22.9%	1:8; 1:11; 8:9; 9:11	.019
Critical thinking	68.3	19.6	77.1	17.0	23.6	1598	< 0.001	0.48	72.9%	4.2%	23.0%	5:10; 5:11	.020
Perseverance	67.3	20.1	76.2	17.3	22.4	1598	< 0.001	0.47	72.4%	4.8%	22.8%	1:11; 6:10	.018
Relationships with adults	64.2	19.7	71.7	18.4	20.7	1598	< 0.001	0.39	71.0%	4.9%	24.1%	1:10; 1:11; 6:10; 6:11	.025
Relationships with peers	72.9	18.2	78.4	16.2	16.8	1598	< 0.001	0.32	64.5%	5.3%	30.2%	8:9; 9:11	.017

Note: Each survey scale ranged from 0 to 99, with a score of 49 representing the midpoint. There were statistically significant differences between states for all nine CIS-S outcomes (all p’s < 0.001 and all n’s = 1599, see main text). For between-group differences column for the states, the asterisk (*) denotes the state with the greater difference between the retrospective posttest and retrospective pretest based on Tukey’s HSD (i.e., State 1*:State 2 indicates that State 1 showed a significantly greater change relative to State 2, whereas State 1:State2* indicates that State 2 showed a greater change relative to State 1)