Online Appendix
Online Appendix Supplement to Figure 3. Proportion of samples per cell with any parcel allocation variability (PAV) in model ranking vs. average LRT statistic per cell. Notes. In each panel, top curve = low communalities; middle curve = medium communalities; bottom curve = high communalities. The 7 dots per curve connected by a spline and represent 7 sizes of structural differences between Models a and b, ranging from small to large. Vertical bar=decision threshold (i.e. critical value of 5.99) for LRT. Proportion of samples with PAV in rank
.45
N=150
.40 .35 .30 .25 .20 .15 .10 .05 0 0
10
20
30
40
50
60
Proportion of samples with PAV in rank
.45
N=250
.40 .35 .30 .25 .20 .15 .10 .05 0 0
10
20
30
40
50
60
Proportion of samples with PAV in rank
.45
N=350
.40 .35 .30 .25 .20 .15 .10 .05 0 0
10
20
30
40
50
60
N=450
Online Appendix Supplement to Figure 4. Proportion of samples per cell with any parcel allocation variability (PAV) in model ranking vs. average ∆BIC per cell. Notes. In each panel, top curve = low communalities; middle curve = medium communalities; bottom curve = high communalities. The 7 dots per curve connected by a spline and represent 7 sizes of structural differences between Models a and b, ranging from small to large. Vertical bar=decision threshold for ∆BIC (i.e., 0). N=150
N=250
N=350
N=450
Online Appendix Supplement to Figure 5. Proportion of samples per cell with any parcel allocation variability (PAV) in model ranking vs. average ∆AIC per cell. Notes. In each panel, top curve = low communalities; middle curve = medium communalities; bottom curve = high communalities. The 7 dots per curve connected by a spline and represent 7 sizes of structural differences between Models a and b, ranging from small to large. Vertical bar=decision threshold for ∆AIC (i.e., 0). N=150
N=250
N=350
N=450
Online Appendix Supplement to Figure 6. Magnitude of parcel-allocation variability (PAV) in model ranking. Notes. Each X = a sample where PAV in model ranking occurred. Magnitude of PAV in rank = proportion of allocations/sample preferring the least-favorable model; maximum magnitude =.50. Vertical bar= decision threshold. LRT=likelihood ratio test. BIC=Bayesian Information Criteria. AIC=Akaike’s Information Criteria. 0.5
0.5
N=150
0.5
N=150
0.4
0.4
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0.0
1
2
3
4
5
6
7
8
9
10
0.5
11
12
13
0.0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
0.5
N=250
6
7
8
N=250
0.0 -4 0.5
0.4
0.4
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0.0
1
2
3
4
5
6
7
8
9
10
11
12
13
0.5
0.0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
0.5
N=350
0.0 -4 0.5
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
2
3
4
5
6
7
8
9
10
11
12
N=450
-2
-1
0
1
2
3
4
13
0.0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
N=450
5
6
N=250
-3
-2
-1
0
1
2
3
4
5
6
N=350
0.4
1
-3
N=350
0.4
0.0
N=150
8
0.0 -4
-3
-2
-1
0
1
2
3
4
5
N=450
6
.45
N=150
.40 .35 .30 .25 .20 .15 .10 .05 0 0
10
20
30
40
50
60
Proportion of samples with PAV in rank
.45
N=250
.40 .35 .30 .25 .20 .15 .10 .05 0 0
10
20
30
40
50
60
Proportion of samples with PAV in rank
.45
N=350
.40 .35 .30 .25 .20 .15 .10 .05 0 0
10
20
30
40
50
60
N=450
Online Appendix Supplement to Figure 4. Proportion of samples per cell with any parcel allocation variability (PAV) in model ranking vs. average ∆BIC per cell. Notes. In each panel, top curve = low communalities; middle curve = medium communalities; bottom curve = high communalities. The 7 dots per curve connected by a spline and represent 7 sizes of structural differences between Models a and b, ranging from small to large. Vertical bar=decision threshold for ∆BIC (i.e., 0). N=150
N=250
N=350
N=450
Online Appendix Supplement to Figure 5. Proportion of samples per cell with any parcel allocation variability (PAV) in model ranking vs. average ∆AIC per cell. Notes. In each panel, top curve = low communalities; middle curve = medium communalities; bottom curve = high communalities. The 7 dots per curve connected by a spline and represent 7 sizes of structural differences between Models a and b, ranging from small to large. Vertical bar=decision threshold for ∆AIC (i.e., 0). N=150
N=250
N=350
N=450
Online Appendix Supplement to Figure 6. Magnitude of parcel-allocation variability (PAV) in model ranking. Notes. Each X = a sample where PAV in model ranking occurred. Magnitude of PAV in rank = proportion of allocations/sample preferring the least-favorable model; maximum magnitude =.50. Vertical bar= decision threshold. LRT=likelihood ratio test. BIC=Bayesian Information Criteria. AIC=Akaike’s Information Criteria. 0.5
0.5
N=150
0.5
N=150
0.4
0.4
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0.0
1
2
3
4
5
6
7
8
9
10
0.5
11
12
13
0.0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
0.5
N=250
6
7
8
N=250
0.0 -4 0.5
0.4
0.4
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0.0
1
2
3
4
5
6
7
8
9
10
11
12
13
0.5
0.0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
0.5
N=350
0.0 -4 0.5
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
2
3
4
5
6
7
8
9
10
11
12
N=450
-2
-1
0
1
2
3
4
13
0.0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
N=450
5
6
N=250
-3
-2
-1
0
1
2
3
4
5
6
N=350
0.4
1
-3
N=350
0.4
0.0
N=150
8
0.0 -4
-3
-2
-1
0
1
2
3
4
5
N=450
6
