Skip to main content
International Journal of STEM Education

Table 1 Overall effect size estimate by study

From: A meta-analytic investigation of the impact of middle school STEM education: where are all the students of color?

Study

# Of effect estimates

Overall effect

SE

Lower

Upper

p

Adams

7

− 0.30

0.17

− 0.63

0.04

0.085

Bentley

2

− 0.50

0.10

− 0.69

− 0.31

0.000

Chine

2

0.70

0.10

0.51

0.89

0.000

Gazibeyoglu & Aydin

1

0.68

0.29

0.12

1.24

0.017

Gerlach

4

0.79

0.36

0.08

1.50

0.028

Gülen & Yaman

1

0.72

0.33

0.08

1.36

0.028

Harlan et al.

3

0.31

1.21

− 2.06

2.68

0.798

Hidge et al.

1

− 0.81

0.31

− 1.43

− 0.20

0.010

Izgi & Kalaysci

1

0.99

0.30

0.40

1.58

0.001

James

4

0.32

0.27

− 0.22

0.85

0.247

Kurt et al.

1

1.04

0.42

0.22

1.86

0.013

Kutch

3

0.06

0.09

− 0.11

0.23

0.486

Moreno et al.

1

1.17

0.13

0.91

1.43

0.000

Olivarez

3

0.59

0.10

0.40

0.79

0.000

Özcan & Koca

1

0.72

0.37

0.00

1.44

0.049

Price et al.

2

0.87

0.20

0.48

1.25

0.000

Sarican & Akgunduz

1

0.22

0.30

− 0.37

0.82

0.457

Selcen Guzey et al.

1

0.23

0.05

0.14

0.32

0.000

Tolliver

2

0.19

0.56

− 0.90

1.28

0.733

Wade-Shepard

4

2.98

1.33

0.38

5.59

0.025

  1. Robust variance estimates (RVE) were not computed because of the limited number of studies in the current investigation that are used for multiple effect size estimates. Hedges et al. (2010) simulations suggest that RVE are valid when there are a larger number of studies with multiple estimators, suggesting twenty or more. The current investigation includes nine studies that produce a single effect size estimation, leaving eleven students with potentially non-independent estimates. Data are broken down by study and by the various moderators, in order to provide the estimates for all possible levels of effect
Back to article page