bayesian evaluation of informative hypotheses in sem using mplus rens van de schoot...

39
Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot [email protected] rensvandeschoot.wordpress.com

Upload: chester-black

Post on 19-Jan-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Bayesian Evaluation of Informative Hypotheses in SEM using Mplus

Rens van de [email protected]

Page 2: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Informative hypotheses

Page 3: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Null hypothesis testing

Difficult to evaluate specific expectations using classical null hypothesis testing:

– Not always interested in null hypothesis

– ‘accepting’ alternative hypothesis no answer

– No direct relation

– Visual inspection

– Contradictory results

Page 4: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Null hypothesis testing Theory

Expectations

Testing:

– H0: nothing is going on

vs.

– H1: something is going on, but we do not know what…

= catch-all hypothesis

Page 5: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Evaluating Informative Hypotheses Theory

Expectations

Evaluating informative hypotheses:

- Ha: theory/expectation 1

vs.

- Hb: theory/expectation 2

vs.

- Hc: theory/expectation 3

etc.

√√

Page 6: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Informative Hypotheses

Hypothesized order constraints between

statistical parameters

Order constraints: < > Statistical parameters: means, regression

coefficients, etc.

Page 7: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Why???

Direct support for your expectation

Gain in power Van de Schoot & Strohmeier, (2011), Testing

informative hypotheses in SEM Increases Power. IJBD vol. 35 no. 2 180-190

7

Page 8: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Default Bayes factors

Page 9: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Default Bayes factors

Page 10: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Default Bayes factors

Page 11: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Bayes factors for informative hypo’s

As was shown by Klugkist et al. (2005, Psych.Met.,10,

477-493), the Bayes factor (BF) of HA versus Hunc can be written as

where fi can be interpreted as a measure for model fit and ci as a measure for model complexity of Ha.

, =,

i

iuncA c

fBF

Page 12: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Bayes factors for informative hypo’s

Model Complexity, ci :– Can be computed before observing

any data. – Determining the number of

restrictions imposed on the means

– The more restriction, the lower ci

Page 13: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Bayes factors for informative hypo’s

Model fit, fi :– After observing some data, – It quantifies the amount of

agreement of the sample means with the restrictions imposed

Page 14: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Bayes factors for informative hypo’s

Bayesian Evaluation of Informative Hypotheses in SEM using Mplus

– Van de Schoot, Hoijtink, Hallquist, & Boelen (in press). Bayesian Evaluation of inequality-constrained Hypotheses in SEM Models using Mplus. Structural Equation Modeling

– Van de Schoot, Verhoeven & Hoijtink (under review). Bayesian Evaluation of Informative Hypotheses in SEM using Mplus: A Black Bear story.

Page 15: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Example: Depression

15

Page 16: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Data

(1) females with a high score on negative coping strategies (n = 1429),

(2) females with a low score on negative coping strategies (n = 1532),

(3) males with a high score on negative coping strategies (n = 1545),

(4) males with a low score on negative coping strategies (n = 1072),

16

Page 17: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Model

17

*40./*41./*47./44. *

Experienced a negative life

event

Depression Time 1

Depression Time 2

001./*04./03./08. *

*19./*18./*22./13. *

*73./*63./*71./61. *

84./83./77./78.

Page 18: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Expectations

“We expected that the relation between life events on Time 1 is a stronger predictor of depression on Time 2 for girls who have a negative coping strategy than for girls with a less negative coping strategy and that the same holds for boys. Moreover, we expected that this relation is stronger for girls with a negative coping style compared to boys with a negative coping style and that the same holds for girls with a less negative coping style compared to boys with a less negative copings style.”

18

Page 19: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Expectations

Hi1 : (β1 > β2) & (β3 > β4)

Hi2 : β1 > (β2, β3) > β4)

19

Page 20: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Model

20

*40./*41./*47./44. *

Experienced a negative life

event

Depression Time 1

Depression Time 2

001./*04./03./08. *

*19./*18./*22./13. *

*73./*63./*71./61. *

84./83./77./78.

Page 21: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Bayes Factor

21

i

i

c

fBF Hu vs.Hi

Hu vs.Hi2

Hu vs.Hi1Hi2 vs.Hi1 BF

BFBF

Page 22: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Step-by-step

22

we need to obtain estimates for fi and ci

Step 1. The first step is to formulate an inequality constrained hypothesis

Step 2. The second step is to compute ci. For simple order restricted hypotheses this can be done by hand.

Page 23: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Step-by-step

23

Count the number of parameters in the inequality constrained hypothesis – in our example: 4 (β1 β2 β3 β4)

Order these parameters in all possible ways: – in our example there are 4! = 4x3x2x1= 24

different ways of ordering four parameters.

Page 24: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Step-by-step

24

Count the number of possible orderings that are in line with each of the informative hypotheses:

– For Hi1 (β1 > β2) & (β3 > β4) that are 6 possibilities;

– For Hi2 β1 > (β2, β3) > β4) that are 2 possibilities;

Page 25: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Step-by-step

25

Divide the value obtained in step 3 by the value obtained in step 2:

– c i1 = 6/24 = 0.25

– c i2 = 2/24 = 0.0833

Note that Hi2 is the most specific hypothesis and receives the smallest value for complexity.

Page 26: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Step-by-step

26

Step 3. Run the model in Mplus:

Page 27: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Mplus syntax

DATA: FILE = data.dat;VARIABLE:NAMES ARE lif1 depr1 depr2 groups;MISSING ARE ALL (-9999);

KNOWNCLASS is g(group = 1 group = 2 group = 3 group = 4);

CLASSES is g(4);27

Page 28: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Mplus syntax

ANALYSIS:TYPE is mixture;

ESTIMATOR = Bayes; PROCESSOR= 32;

28

Page 29: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Mplus syntax

MODEL:%overall%depr2 on lif1;depr2 on depr1;lif1 with depr1;[lif1 depr1 depr2]; lif1 depr1 depr2; 

29

Page 30: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Mplus syntax

 !save the parameter estimates for each iteration:

SAVEDATA: BPARAMETERS are

c:/Bayesian_results.dat;

30

Page 31: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

31

Page 32: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Using MplusAutomation

Page 33: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

R syntax

To install MplusAutomation:

R: install.packages(c("MplusAutomation"))R: library(MplusAutomation)

Specify directory:R: setwd("c:/mplus_output")

33

Page 34: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

R syntax

Locate output file of Mplus: R: btest <- getSavedata_Bparams("output.out")

Compute f1:

R: testBParamCompoundConstraint (btest, "( STDYX_.G.1...DEPR2.ON.LIF_1 > STDYX_.G.2...DEPR2.ON.LIF_1) & STDYX_.G.3...DEPR2.ON.LIF_1 > TDYX_.G.4...DEPR2.ON.LIF_1)")

34

Page 35: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

R syntax

Compute f2:

R: testBParamCompoundConstraint(btest, "( STDYX_.G.1...DEPR2.ON.LIF_1 > STDYX_.G.2...DEPR2.ON.LIF_1) & (STDYX_.G.3...DEPR2.ON.LIF_1 > STDYX_.G.4...DEPR2.ON.LIF_1)& (STDYX_.G.1...DEPR2.ON.LIF_1 > STDYX_.G.3...DEPR2.ON.LIF_1)& STDYX_.G.2...DEPR2.ON.LIF_1 > STDYX_.G.4...DEPR2.ON.LIF_1)") 

35

Page 36: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Results

 fi1 = .7573

c i1 = 0.25

fi2 = .5146

c i2 = 0.0833

36

Page 37: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Results

 BF1 vs unc = .7573 / .25 = 3.03

BF2 vs unc = .5146 / .0833 = 6.18

37

Page 38: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Results

 BF1 vs unc = .7573 / .25 = 3.03

BF2 vs unc = .5146 / .0833 = 6.18

BF 2 vs 1 = 6.18 / 3.03 = 2.04

38

Page 39: Bayesian Evaluation of Informative Hypotheses in SEM using Mplus Rens van de Schoot a.g.j.vandeschoot@uu.nl rensvandeschoot.wordpress.com

Conclusions

Excellent tool to include prior knowledge if available

Direct support for you expectations!

Gain in power