Different evolution acceptance instruments lead to different research findings

Table 5 Evaluation of confidence intervals of unstandardized beta coefficients from predictor variables across evolution acceptance instruments used to determine if results across analyses are different

	MATE/Micro	MATE/Macro	MATE/Human	MATE/GAENE	MATE/100	Micro/Macro	Micro/Human	Micro/GAENE	Micro/100	Macro/Human	Macro/GAENE	Macro/100	Human/GAENE	Human/100	GAENE/100	# of diff
College GPA	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	0
Course level	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	0
Parent education	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	X	✔	X	✔	2
Evolution understanding	✔	X	X	X	X	X	✔	X	X	✔	X	✔	X	✔	✔	9
NOS understanding	✔	X	✔	X	✔	X	✔	X	✔	✔	✔	✔	✔	✔	✔	4
Religiosity	✔	✔	✔	✔	X	✔	✔	✔	X	✔	✔	X	✔	✔	X	4
Catholic	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	0
Protestant/LDS	X	✔	X	✔	X	X	X	✔	X	✔	✔	X	✔	X	X	9
Muslim	✔	✔	✔	✔	✔	✔	X	✔	✔	✔	✔	✔	X	✔	✔	2
Other religion	X	X	✔	X	X	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	4
Democrat	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	0
Republican	✔	✔	✔	✔	X	✔	✔	✔	X	✔	✔	X	✔	X	X	5
Black/Af. Am.	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	0
Hispanic	✔	✔	✔	✔	X	✔	X	✔	X	✔	✔	X	✔	✔	✔	4
Asian	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	0
Other race/ethnicity	✔	✔	X	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔	3

Each column shows the comparison between two evolution acceptance instruments. Each row is a predictor variable that was included in every regression. Each cell indicates whether the confidence intervals of two instruments overlap. If the confidence intervals overlap, it is indicated with a “✔”; a “X” indicates that the two confidence intervals do not overlap and thus the beta coefficients are different from one another beyond sampling variability. The far right column summarizes the number of instances where confidence intervals do not overlap. Note that this is a conservative approach and those with overlapped confidence intervals may also be different (Schenker and Gentleman 2001)

ISSN: 1936-6434