* Set up your data file such that coder A's F.Ex codes are arranged in columns 1 through 3,
* one column for each F.Ex digit (e.g., 3 | 1 | 1).
* This also means that each F.Ex code occupies one row (no matter whether some behaviors had multiple explanations).
* Coder B's F.Ex codes are similarly arranged and occupy columns 4 through 6.
** Transform first F.Ex code into codes for explanation modes.
RECODE A1 B1 (0 = 0)(1 thru 6 = 1) into A_code B_code.
RECODE A1 B1 (0 = sysmis)(1 = 1)(2 = 2)(3,4 = 3) (6 = 6) into A_mode B_mode .
RECODE A1 B1 (2 = 2)(3,4 = 3)(0, 1, 6 = sysmis) into A_mode2 B_mode2.
VALUE LABELS A_mode B_mode A_mode2 B_mode2 1 'cause' 2 'CHR' 3 'reason' 6 'enabfac'.

Reliab For Codeability

CROSSTAB A_code by B_code
/cells = count tot
/sta = kappa gamma.

Warnings
No measures of association are computed for the crosstabulation of A_code * B_code. At least one variable in each 2-way table upon which measures of association are computed is a constant.

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_code * B_code 253 100.0% 0 .0% 253 100.0%

A_code * B_code Crosstabulation
  B_code
1 Total
A_code 1 Count 253 253
% of Total 100.0% 100.0%
Total Count 253 253
% of Total 100.0% 100.0%

Symmetric Measures
  Value
Ordinal by Ordinal Gamma .a
Measure of Agreement Kappa .
N of Valid Cases 253
a. No statistics are computed because A_code and B_code are constants.


Reliab For all EXPLANATION MODES

CROSSTAB A_mode by B_mode
/cells = count tot
/sta = kappa gamma.

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_mode * B_mode 253 100.0% 0 .0% 253 100.0%

A_mode * B_mode Crosstabulation
  B_mode
cause CHR reason Total
A_mode cause Count 34 0 0 34
% of Total 13.4% .0% .0% 13.4%
CHR Count 0 38 4 42
% of Total .0% 15.0% 1.6% 16.6%
reason Count 0 18 159 177
% of Total .0% 7.1% 62.8% 70.0%
Total Count 34 56 163 253
% of Total 13.4% 22.1% 64.4% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma .989 .006 13.143 .000
Measure of Agreement Kappa .824 .036 17.557 .000
N of Valid Cases 253      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.


Reliab For CHR vs. reason

CROSSTAB A_mode2 by B_mode2
/cells = count tot
/sta = kappa gamma.

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_mode2 * B_mode2 219 86.6% 34 13.4% 253 100.0%

A_mode2 * B_mode2 Crosstabulation
  B_mode2
CHR reason Total
A_mode2 CHR Count 38 4 42
% of Total 17.4% 1.8% 19.2%
reason Count 18 159 177
% of Total 8.2% 72.6% 80.8%
Total Count 56 163 219
% of Total 25.6% 74.4% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma .976 .014 8.118 .000
Measure of Agreement Kappa .712 .057 10.725 .000
N of Valid Cases 219      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.


** Transform third F.Ex code of reason explanations into reason type [1, 2, 3].
RECODE A3 B3 (0 =sysmis)(else = copy) into A_reatyp B_reatyp.
VALUE LABELS A_reatyp B_reatyp 1 'des' 2 'bel' 3 'val'.

Reliab For REASON TYPE

TEMP.
SEL IF (A_mode = 3 and B_mode = 3).
CROSSTAB A_reatyp by B_reatyp
/cells = count tot
/sta = kappa gamma.

 

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_reatyp * B_reatyp 159 100.0% 0 .0% 159 100.0%

A_reatyp * B_reatyp Crosstabulation
  B_reatyp
des bel val Total
A_reatyp des Count 89 1 1 91
% of Total 56.0% .6% .6% 57.2%
bel Count 4 53 4 61
% of Total 2.5% 33.3% 2.5% 38.4%
val Count 0 1 6 7
% of Total .0% .6% 3.8% 4.4%
Total Count 93 55 11 159
% of Total 58.5% 34.6% 6.9% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma .978 .019 18.562 .000
Measure of Agreement Kappa .869 .037 12.747 .000
N of Valid Cases 159      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.


** Analyze first F.Ex code of reason explanations as marked vs. unmarked reasons.
RECODE A1 B1 (3 = 1) (4 = 0) (else = sysmis) into A_mark B_mark.
VALUE LABELS A_mark B_mark 1 'marked' 0 'unmarked'

Reliab ForMENTAL STATE MARKERS

TEMP.
SEL IF (A_mode = 3 and B_mode = 3).
CROSSTAB A_mark by B_mark
/cells = count tot
/sta = kappa gamma.

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_mark * B_mark 159 100.0% 0 .0% 159 100.0%

A_mark * B_mark Crosstabulation
  B_mark
unmarked marked Total
A_mark unmarked Count 77 10 87
% of Total 48.4% 6.3% 54.7%
marked Count 8 64 72
% of Total 5.0% 40.3% 45.3%
Total Count 85 74 159
% of Total 53.5% 46.5% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma .968 .016 14.960 .000
Measure of Agreement Kappa .772 .051 9.739 .000
N of Valid Cases 159      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.


** Transform second F.Ex code into either person-situation-interaction (PS3) or just
** person/interaction vs. situation (PS2).
RECODE A2 B2 (0 = sysmis) (1=1)(2,4,6 = 2)(3,5,7 = 3) into A_PS3 B_PS3.
RECODE A2 B2 (0 = sysmis) (1,3,5,7 = 1) (2,4,6 = 2) into A_PS2 B_PS2.
VALUE LABELS A_PS3 B_PS3 1 'person' 2 'situation' 3 'interaction'
/A_PS2 B_PS2 1 'pers_ia' 2 = sit'.
** For reason explanations, these PS codes correspond to reason contents.

Reliab For REASON CONTENT

TEMP.
SEL IF (A_mode = 3 and B_mode = 3).
CROSSTAB A_PS3 by B_PS3
/cells = count tot
/sta = kappa gamma.

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_PS3 * B_PS3 159 100.0% 0 .0% 159 100.0%

A_PS3 * B_PS3 Crosstabulation
  B_PS3
person situation interaction Total
A_PS3 person Count 88 1 11 100
% of Total 55.3% .6% 6.9% 62.9%
situation Count 4 17 9 30
% of Total 2.5% 10.7% 5.7% 18.9%
interaction Count 6 0 23 29
% of Total 3.8% .0% 14.5% 18.2%
Total Count 98 18 43 159
% of Total 61.6% 11.3% 27.0% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma .825 .052 8.902 .000
Measure of Agreement Kappa .640 .055 11.037 .000
N of Valid Cases 159      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.


** For non-reason explanations, the PS codes refer to cause types.

Reliab For CAUSE TYPE

TEMP.
SEL IF (A_mode LE 2 and B_mode LE 2).
CROSSTAB A_PS3 by B_PS3
/cells = count tot
/sta = kappa gamma.

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_PS3 * B_PS3 72 100.0% 0 .0% 72 100.0%

A_PS3 * B_PS3 Crosstabulation
  B_PS3
person situation interaction Total
A_PS3 person Count 44 2 4 50
% of Total 61.1% 2.8% 5.6% 69.4%
situation Count 0 4 1 5
% of Total .0% 5.6% 1.4% 6.9%
interaction Count 4 0 13 17
% of Total 5.6% .0% 18.1% 23.6%
Total Count 48 6 18 72
% of Total 66.7% 8.3% 25.0% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma .871 .065 5.421 .000
Measure of Agreement Kappa .676 .088 7.079 .000
N of Valid Cases 72      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.



** Treat third F.Ex code in non-reason explanations as specific person cause (1 = behavior, 2 = internal state, etc.).

DO IF (A_mode NE 3 and B_mode NE 3).
COMP A_specP = A3.
COMP B_specP = B3.
END IF.
VALUE LABELS A_specP B_specP 1 'beh' 2 'internal' 3 'attperc' 4 'pro_att' 5 'trait' 6 'pass' 7 'stable_pro' 8 'group'.

** Or turn these specific codes into one broad distinction -- between trait causes and nontrait causes.
RECODE A_specP B_specP (0,1,2,3,4,7,6,8 = 0)(5,9 = 1) into A_trait B_trait.
VALUE LABELS A_trait B_trait 0 'nontrait' 1 'trait'.

Reliab For PERSON CAUSE TYPE


TEMP.
SEL IF (A_mode LE 2 and B_mode LE 2).
CROSSTAB A_specP by B_specP / A_trait by B_trait
/cells = count tot
/sta = kappa gamma.

Case Processing Summary
  Cases
Valid Missing Total
N Percent N Percent N Percent
A_specP * B_specP 72 100.0% 0 .0% 72 100.0%
A_trait * B_trait 72 100.0% 0 .0% 72 100.0%


A_specP * B_specP

Crosstab
  B_specP
beh internal attperc pro_att trait stable_pro group Total
A_specP beh Count 12 0 0 1 0 0 0 13
% of Total 16.7% .0% .0% 1.4% .0% .0% .0% 18.1%
internal Count 1 15 0 1 1 0 0 18
% of Total 1.4% 20.8% .0% 1.4% 1.4% .0% .0% 25.0%
attperc Count 0 1 2 3 0 0 0 6
% of Total .0% 1.4% 2.8% 4.2% .0% .0% .0% 8.3%
pro_att Count 2 1 1 18 0 0 0 22
% of Total 2.8% 1.4% 1.4% 25.0% .0% .0% .0% 30.6%
trait Count 0 0 0 0 8 0 0 8
% of Total .0% .0% .0% .0% 11.1% .0% .0% 11.1%
pass Count 1 0 0 0 0 0 0 1
% of Total 1.4% .0% .0% .0% .0% .0% .0% 1.4%
stable_pro Count 0 0 0 0 1 1 0 2
% of Total .0% .0% .0% .0% 1.4% 1.4% .0% 2.8%
group Count 0 0 0 1 0 0 1 2
% of Total .0% .0% .0% 1.4% .0% .0% 1.4% 2.8%
Total Count 16 17 3 24 10 1 1 72
% of Total 22.2% 23.6% 4.2% 33.3% 13.9% 1.4% 1.4% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma .813 .077 8.835 .000
Measure of Agreement Kappa .c      
N of Valid Cases 72      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.
c. Kappa statistics cannot be computed.They require a symmetric 2-way table in which the values of the first variable match the values of the second variable.


A_trait * B_trait

Crosstab
  B_trait
nontrait trait Total
A_trait nontrait Count 62 2 64
% of Total 86.1% 2.8% 88.9%
trait Count 0 8 8
% of Total .0% 11.1% 11.1%
Total Count 62 10 72
% of Total 86.1% 13.9% 100.0%

Symmetric Measures
  Value Asymp. Std. Errora Approx. Tb Approx. Sig.
Ordinal by Ordinal Gamma 1.000 .000 3.418 .001
Measure of Agreement Kappa .873 .088 7.470 .000
N of Valid Cases 72      
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.