Skip to main content

Table 3 Measurement invariance tests across age, race, sex, income, sexual identity, and sexual orientation groups

From: Psychometric evaluation of an Adverse Childhood Experiences (ACEs) measurement tool: an equitable assessment or reinforcing biases?

Modela

Tests of Invariance & Structure

df

CFI

TLI

RMSEA [90% C.I.]

SRMR

Δ df

Δ CFI

Δ TLI

A. Across Age Groups

  A1

Measurement model

246

0.994

0.992

.017 [.016 − .018]

.044

--

--

--

  A2 a

Second-order model/configural invariance

258

0.988

0.985

.024 [.023 − .025]

.062

12

.006

.007

  A3 b

First-order within-scale parallel model

306

0.984

0.983

.026 [.025 − .026]

.073

48

.004

.002

  A4

First- and second-order metric invariance

321

0.984

0.984

.025 [.024 − .026]

.075

15

.000

+ .001

  A5

First- and second-order scalar invariance

401

0.981

0.984

.024 [.024 − .025]

.078

80

.003

.000

B. Across Race Groups

  B1

Measurement model

205

0.994

0.992

.015 [.014 − .016]

.037

--

--

--

  B2 a

Second-order model/configural invariance

215

0.986

0.982

.023 [.022 − .023]

.057

10

.008

.010

  B3 b

First-order within-scale parallel model

255

0.983

0.982

.022 [.022 − .023]

.069

40

.003

+ .000

  B4

First- and second-order metric invariance

267

0.984

0.984

.022 [.021 − .022]

.070

12

+ .001

+ .002

  B5

First- and second-order scalar invariance

331

0.983

0.986

.020 [.019 − .021]

.071

64

.001

+ .002

 C. Across Sex Groups

  C1

Measurement model

82

0.994

0.992

.016 [.015 − .017]

.038

--

--

--

  C2 a

Second-order model/configural invariance

86

0.987

0.984

.023 [.022 − .023]

.059

4

.007

.008

  C3 b

First-order within-scale parallel model

102

0.985

0.984

.022 [.021 − .023]

.069

16

.002

+ .000

  C4

First- and second-order metric invariance

105

0.985

0.985

.022 [.021 − .022]

.070

3

+ .000

+ .001

  C5

First- and second-order scalar invariance

121

0.984

0.986

.021 [.020 − .022]

.071

16

.001

+ .001

D. Across Income (SES) Groups

  D1

Measurement model

205

0.993

0.991

.018 [.017 − .019]

.040

--

--

--

  D2 a

Second-order model/configural invariance

215

0.984

0.980

.026 [.025 − .027]

.062

10

.009

.011

  D3 b

First-order within-scale parallel model

255

0.980

0.979

.027 [.026 − .028]

.074

40

.004

.001

  D4

First- and second-order metric invariance

267

0.981

0.981

.026 [.025 − .026]

.074

12

+ .001

+ .002

  D5

First- and second-order scalar invariance

331

0.981

0.984

.023 [.022 − .024]

.075

64

.000

+ .003

E. Sexual Identity Groups

  E1

Measurement model

82

0.995

0.993

.013 [.012 − .014]

.037

--

--

--

  E2 a

Second-order model/configural invariance

86

0.988

0.984

.019 [.018 − .020]

.057

4

.007

.009

  E3 b

First-order within-scale parallel model

102

0.987

0.986

.018 [.017 − .019]

.068

16

.001

+ .002

  E4

First- and second-order metric invariance

105

0.988

0.987

.018 [.017 − .019]

.068

3

+ .001

+ .001

  E5

First- and second-order scalar invariance

121

0.988

0.989

.016 [.015 − .017]

.068

16

.000

+ .002

 F. Sexual Orientation Groups

  F1

Measurement model

82

0.996

0.994

.013 [.012 − .015]

.039

--

--

--

  F2 a

Second-order model/configural invariance

90

0.990

0.987

.020 [.019 − .021]

.061

8

.006

.007

  F3 b

First-order within-scale parallel model

102

0.988

0.987

.021 [.020 − .022]

.069

12

.002

.000

  F4

First- and second-order metric invariance

105

0.987

0.987

.021 [.020 − .022]

.070

3

.001

.000

  F5

First- and second-order scalar invariance

121

0.988

0.989

.019 [.018 − .020]

.070

16

+ .001

+ .002

G. Common Factor Model

  G1

Measurement model

41

0.993

0.991

.017 [.016 − .017]

.036

--

--

--

  G2 a

Second-order model

43

0.987

0.983

.023 [.022 − .023]

.056

2

.006

.008

  G3 b

First-order within-scale parallel model

51

0.986

0.985

.021 [.020 − .022]

.066

8

.001

+ .002

  1. aFirst-order factor loadings were set to be equal within groups to obtain an over-identified model
  2. bFirst-order within-factor items are constrained to be equal but allowed to vary across groups