Error preventing item fit from being calculated for Multiple Group Estimation model

[Daniel McKelvey]

Hi Phil,

I've encountered a new error for me that is preventing me from calculating item fit for an MGE that I am using to test for DIF:

> HetvGB_config_itemfit = itemfit(HetvGB_config)

Error in scores[match(sfulldata, stabdata2), , drop = FALSE] : 

  subscript out of bounds

The instrument is a 21 item measure with 7 response categories per item. I've listwise deleted any participant with missing data, so that should not be the issue. I also modeled each group separately and calculated item fit that way with no errors, so I'm not sure what has happened.

Thanks for your time,

Daniel

[Phil Chalmers]

This one looks less familiar to me. Could you send me simple reproducible code and data?

Phil

--
You received this message because you are subscribed to the Google Groups "mirt-package" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mirt-package...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Daniel McKelvey]

Sure! Here is my code:

##Group A Data

M_GrpA = read.csv("M_GRPA.csv", header = T)

M_GrpA_Items = M_GrpA[ ,  c(6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26)]

M_GrpA_Items = na.omit(M_GrpA_Items)

##Group B Data

M_GrpB = read.csv("M_GRPB.csv", header = T)

M_GrpB_Items = M_GrpB[ , c(6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26)]

M_GrpB_Items = na.omit(M_GrpB_Items)

##Binding retrieved data into single file

M_GrpAvB_items = rbind(M_GrpA_Items, M_GrpB_Items)

groups = c(rep('A',1820), rep('B', 481))

model ='TMI = 1-21'

#Configural invariance model

AvB_config = multipleGroup(M_GrpAvB_items, model, group = groups,

                               technical = list(NCYCLES=5000))

itemfit(AvB_config)

Thank you for your help!

[Phil Chalmers]

Thanks. Must be a patched bug already. I get this as the output:

> itemfit(AvB_config)

$A

   item    S_X2 df.S_X2 p.S_X2

1    M1 281.902     287  0.574

2    M2 243.500     213  0.074

3    M3 216.339     186  0.063

4    M4 293.353     269  0.147

5    M5 182.969     178  0.384

6    M6 368.740     329  0.065

7    M7 380.132     328  0.025

8    M8 280.581     251  0.097

9    M9 334.016     298  0.074

10  M10 322.960     255  0.003

11  M11 267.844     255  0.278

12  M12 195.113     171  0.100

13  M13 280.260     280  0.484

14  M14 438.646     390  0.045

15  M15 249.173     230  0.184

16  M16 345.818     299  0.032

17  M17 483.532     399  0.002

18  M18 268.736     291  0.821

19  M19 382.624     351  0.118

20  M20 357.672     333  0.169

21  M21 352.490     299  0.018

$B

   item    S_X2 df.S_X2 p.S_X2

1    M1  60.585      61  0.491

2    M2  71.686      51  0.030

3    M3  93.884      57  0.002

4    M4  85.268      70  0.103

5    M5  57.735      47  0.136

6    M6 153.409     130  0.079

7    M7 112.257     132  0.893

8    M8 112.839      88  0.038

9    M9 104.579     100  0.357

10  M10  65.712      88  0.964

11  M11 103.360      90  0.159

12  M12  83.759      60  0.023

13  M13 115.591      80  0.006

14  M14 162.109     145  0.157

15  M15  71.247      76  0.633

16  M16 107.948     110  0.538

17  M17 166.898     155  0.243

18  M18 134.316     117  0.131

19  M19 156.265     141  0.179

20  M20 140.450     142  0.521

21  M21 106.344     113  0.658

I'm submitting a new version of mirt tonight, so you should see an updated version on CRAN soon.

Phil