Docstoc

ICPSRweek3Class2Supplement

Document Sample
ICPSRweek3Class2Supplement Powered By Docstoc
					Three Group Model Examples
ThreeGroup1A.ls8        LY=IN GA=IN Chi-square = 1061.5 df=242
                                  IFI =.968 RMSEA = .0511
ThreeGroup1b.ls8        LY=PS GA=IN Chi-square = 985.2 df=230
                                  IFI=.971 RMSEA=.0503
ThreeGroup1c.ls8        LY=IN GA=PS Chi-square=860.13 df=198
                                  IFI=.973 RMSEA=.0503
From ThreeGroup1A:
           United Kingdom:
     Modification Indices for LAMBDA-Y
        RELIG        MORAL
      -------- --------
   A006 10.5092 0.0000
   F028 0.6423 5.6826
   F066 0.0873 5.4233
   F063 7.2921 10.7580
   F118 10.2494 2.0448
   F119 0.4205 0.4225
   F120 1.9053 0.8206
   F121 1.7768 15.9429
Expected Change for LAMBDA-Y
        RELIG        MORAL
      -------- --------
   A006 -0.0417 -0.0001
   F028 -0.0338 -0.0718
   F066 0.0067 -0.0552
   F063 -0.0839 -0.1033
   F118 -0.1595 -0.0770
   F119 0.0291 -0.0236
   F120 0.0602 -0.0346
   F121 0.0513 0.1277
Common (constrained to equality) estimate:
GAMMA
       GENDER           AGE       EDUC TOWNSIZE             OCC1 OCC2
      -------- -------- -------- -------- -------- --------
  RELIG 0.7781 -0.0248 0.0941 0.0474 -0.1304 -0.1015
      (0.0628) (0.0019) (0.0257) (0.0142) (0.1547) (0.1476)
       12.3862 -13.3335 3.6668 3.3387 -0.8431 -0.6878

  MORAL -0.0416 -0.0250 0.3442 0.0582 0.3001 0.4688
      (0.0813) (0.0024) (0.0347) (0.0187) (0.1978) (0.1878)
       -0.5116 -10.2107 9.9237 3.1130 1.5167 2.4958
          OCC3         OCC4
      -------- --------
  RELIG 0.0692 -0.0638
      (0.1317) (0.1336)
        0.5251 -0.4774

  MORAL 0.0425 0.2703
     (0.1635) (0.1666)
      0.2598 1.6221
Britain:
Modification Indices for GAMMA
           GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
         -------- -------- -------- -------- -------- --------
  RELIG 0.2655 0.1056 21.7331 17.4726 14.2093 4.4895
  MORAL 0.0286 3.3527 2.4189 2.1719 8.1511 3.9795

    Modification Indices for GAMMA
         OCC3         OCC4
      -------- --------
  RELIG 7.9096 1.1395
  MORAL 5.2856 0.4618


Expected Change for GAMMA
        GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
      -------- -------- -------- -------- -------- --------
  RELIG -0.1434 0.0013 -0.2001 -0.2424 -0.7350 -0.4939
  MORAL -0.0178 -0.0095 0.0740 -0.0991 0.6175 0.5297

    Expected Change for GAMMA
         OCC3         OCC4
      -------- --------
  RELIG 0.6076 -0.1637
  MORAL -0.4773 0.1092
From model ThreeGroup1C (GA=PS, LY=IN):
USA
GAMMA
        GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
      -------- -------- -------- -------- -------- --------
  RELIG 0.7365 -0.0187 0.1097 0.0227 0.3743 0.2087
      (0.1164) (0.0035) (0.0432) (0.0331) (0.3516) (0.3454)
        6.3293 -5.3459 2.5424 0.6850 1.0644 0.6043

  MORAL 0.1044 -0.0159 0.3039 -0.0257 -0.0657 0.1795
      (0.1563) (0.0048) (0.0595) (0.0452) (0.4801) (0.4718)
        0.6681 -3.3364 5.1058 -0.5700 -0.1368 0.3805
         OCC3         OCC4
      -------- --------
  RELIG 0.5079 0.4624
      (0.3405) (0.3385)
        1.4914 1.3661

  MORAL -0.0523 0.2652
      (0.4648) (0.4621)
       -0.1126 0.5739
CANADA
GAMMA
        GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
      -------- -------- -------- -------- -------- --------
  RELIG 0.8874 -0.0270 0.1301 0.0603 -0.2752 -0.2117
      (0.0903) (0.0027) (0.0440) (0.0165) (0.3395) (0.3140)
        9.8222 -9.8491 2.9542 3.6622 -0.8104 -0.6743

  MORAL -0.0799 -0.0230 0.4065 0.0899 0.2223 0.2770
     (0.1169) (0.0036) (0.0596) (0.0220) (0.4526) (0.4186)
      -0.6832 -6.3710 6.8225 4.0905 0.4911 0.6619

    GAMMA
         OCC3         OCC4
      -------- --------
  RELIG -0.1291 -0.2408
      (0.2956) (0.2955)
       -0.4368 -0.8148

  MORAL -0.0776 0.0202
     (0.3940) (0.3938)
      -0.1969 0.0513
UK
GAMMA
       GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
     -------- -------- -------- -------- -------- --------
 RELIG 0.7531 -0.0280 -0.0634 -0.2261 -0.6057 -0.3037
     (0.1444) (0.0043) (0.0551) (0.0606) (0.2748) (0.2880)
       5.2166 -6.5398 -1.1510 -3.7304 -2.2041 -1.0548

 MORAL -0.1109 -0.0351 0.2670 -0.1420 0.8540 0.9791
    (0.1646) (0.0050) (0.0642) (0.0696) (0.3174) (0.3333)
     -0.6736 -6.9769 4.1593 -2.0403 2.6909 2.9374

   GAMMA
        OCC3         OCC4
     -------- --------
 RELIG 0.2044 -0.1803
     (0.1789) (0.2004)
       1.1428 -0.8996

 MORAL -0.0697 0.4082
    (0.2057) (0.2311)
     -0.3387 1.7662
Common metric completely standardized solution
U.S.
GAMMA
        GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
      -------- -------- -------- -------- -------- --------
  RELIG 0.2008 -0.1763 0.0797 0.0270 0.0607 0.0424
  MORAL 0.0227 -0.1192 0.1763 -0.0245 -0.0085 0.0291:
Canada
GAMMA
        GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
      -------- -------- -------- -------- -------- --------
  RELIG 0.2419 -0.2542 0.0944 0.0718 -0.0446 -0.0430
  MORAL -0.0174 -0.1726 0.2358 0.0856 0.0288 0.0449
UK
GAMMA
        GENDER          AGE        EDUC TOWNSIZE            OCC1 OCC2
      -------- -------- -------- -------- -------- --------
  RELIG 0.2053 -0.2630 -0.0461 -0.2693 -0.0982 -0.0616
  MORAL -0.0241 -0.2635 0.1549 -0.1352 0.1106 0.1587
Various programs for specialized significance tests:

ThreeGroup1d.ls8 LY=IN GA=IN (groups 1= 2 but group ≠3)
In group 3: GA=PS or GA=IN [same result]
     Then: FR GA(1,1)-GA(2,8)                      Chi-square=989.35 df=226
                                       (contrast with: GA=PS 3 groups chi-sq=860.13
                                                               GA=IN 3 groups chi-sq=1061.5)
ThreeGroup1e.ls8         LY=IN GA=IN
             Then FR GA 1 3 in group 3 (test effect of education in UK
                         is different from Canada + US) chi-sq=1039.36 df=241
ThreeGroup1f.ls8         LY=IN GA=PS
             In group 2: EQ GA 1 1 4 GA 2 1 4
             In group 3: FR GA 1 4 (!)
                         Test differences in community size effect on relig,
                         Canada vs. US chi-square = 897.34 df=199
                         (contrast with 1c, chi-square = 860.13 df=198
     (!) even with GA=PS, in group 3 any constraints would be “copied” ; without the FR GA 1 4
     statement in group 3, we’d get the equivalent of:
     EQ GA 1 1 4 GA 2 1 4 GA 3 1 4

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:1
posted:11/21/2013
language:Unknown
pages:7