Root Mean Square Error of Approximation (RMSEA): A measure of how well the factor model reproduces the observed correlations, per degree of freedom. Lower is better. Conventional benchmarks: values at or below .05 indicate good fit, .05 to .08 acceptable fit, and above .10 poor fit. Unlike some indices it rewards simpler models, so it does not automatically improve just by adding factors.

Tucker-Lewis Index (TLI)): A comparative fit index that measures how much better the model fits than a baseline model in which the items are assumed unrelated. It runs roughly from 0 to 1, with higher values indicating better fit: .90 or above is generally considered acceptable and .95 or above good. Values can sit below these benchmarks for long instruments with many small factors, so it is read alongside RMSEA and the residuals rather than in isolation.

Bayesian Information Criterion (BIC): A model comparison measure that balances how well a model fits against how many factors it uses, penalising added complexity. Lower (more negative) values indicate a better balance of fit and parsimony. BIC has no absolute cut-off for “good” or “bad”; it is only meaningful when comparing candidate models fitted to the same data (for example, choosing between 8, 9, and 10 factors).

Exploratory Factor Analysis: JAWS Difference by Distress

expand

Step 1: Sample Size

Total sample: 602

Sample retained (i.e., answered > 50% of the 48 NewDiff items): N = 576 ( 26 excluded for more than 50% missing).

Cases per item (final n ÷ n of items) = 12 . This exceeds the 10:1 ratio commonly used for EFA, so the sample should be adequate for the number of items.

A sample of 300 or more should be sufficient to produce a stable factor solution (Field, pg. 769), though he also noted sampling adequacy is better assessed with KMO.

Note: JAWS item 18 (managing juries) had the lowest response rate (n = 238; 41% of the sample). This is likely because judicial officers in the lower courts made up a significant proportion.

Table 1. Ten NewDiff items with the lowest valid N
Item Valid_N
JAWS_18_NewDiff 238
JAWS_46_NewDiff 382
JAWS_45_NewDiff 423
JAWS_23_NewDiff 458
JAWS_30_NewDiff 482
JAWS_40_NewDiff 486
JAWS_10_NewDiff 493
JAWS_44_NewDiff 502
JAWS_47_NewDiff 503
JAWS_38_NewDiff 520

Step 2: Correlation Matrix

No item pairs correlated above .90 (Table 2).

Determinant = 0.00000000001681289 (below Field’s threshold (> .00001), expected with 48 items as the determinant shrinks with every correlated pair; no pairs above .90, so no items removed on this basis).

43.4 % of item correlations exceeded .30 (mean |r| = 0.29 ; strongest |r| = 0.77 ); the strongest pairs are shown in Table 2, tells us the items share enough common variance to factor analyse.

Table 2. Top 10 strongest item-pair correlations
Item_1 Item_2 r
JAWS_4_NewDiff JAWS_8_NewDiff 0.77
JAWS_22_NewDiff JAWS_24_NewDiff 0.73
JAWS_29_NewDiff JAWS_41_NewDiff 0.72
JAWS_21_NewDiff JAWS_48_NewDiff 0.70
JAWS_45_NewDiff JAWS_46_NewDiff 0.68
JAWS_11_NewDiff JAWS_12_NewDiff 0.65
JAWS_5_NewDiff JAWS_6_NewDiff 0.64
JAWS_37_NewDiff JAWS_38_NewDiff 0.64
JAWS_11_NewDiff JAWS_14_NewDiff 0.63
JAWS_22_NewDiff JAWS_27_NewDiff 0.60

Item 18 (managing juries), correlations with all other items:

Item 18’s mean |r| = 0.19 is well below the overall mean of .29, and even its strongest correlation is only 0.32 (Table 3). Might talk to Richard about dropping this item later.

Table 3. Item 18: 10 strongest correlations with other items
Item r
JAWS_14_NewDiff JAWS_14_NewDiff 0.32
JAWS_45_NewDiff JAWS_45_NewDiff 0.32
JAWS_6_NewDiff JAWS_6_NewDiff 0.28
JAWS_17_NewDiff JAWS_17_NewDiff 0.28
JAWS_19_NewDiff JAWS_19_NewDiff 0.28
JAWS_13_NewDiff JAWS_13_NewDiff 0.27
JAWS_35_NewDiff JAWS_35_NewDiff 0.27
JAWS_44_NewDiff JAWS_44_NewDiff 0.27
JAWS_1_NewDiff JAWS_1_NewDiff 0.26
JAWS_4_NewDiff JAWS_4_NewDiff 0.26

Step 3: Bartlett’s Test of Sphericity

Bartlett’s test of sphericity was significant, chi-square(1128) = 13,847.49, p < .001 (Table 4). The correlation matrix differs from an identity matrix (then all correlation coefficients = 0) and the items are sufficiently interrelated for factor analysis (pg. 775). But this test is almost always significant with large samples, so it should be treated as a minimum requirement rather than evidence of factorability.

Table 4. Bartlett’s test of sphericity
Statistic Value
Chi-square 13,847.49
df 1128
p-value < .001

Step 4: Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy

Overall MSA = 0.93. above the .50 minimum and in the ‘superb’ range (Field, pg. 776). All individual item MSAs exceeded .50 (range 0.79 to 0.97; see Table 5), so no items needed to be excluded. Notably, item 18’s MSA (0.81) was acceptable. Together with the significant Bartlett’s test, the KMO confirms the data were suitable for factor analysis.

Overall MSA = 0.93 | item MSAs ranged from 0.79 to 0.97

Table 5. Per-item KMO (measure of sampling adequacy), sorted lowest to highest
Item MSA
JAWS_46_NewDiff 0.79
JAWS_18_NewDiff 0.81
JAWS_45_NewDiff 0.81
JAWS_32_NewDiff 0.86
JAWS_22_NewDiff 0.90
JAWS_30_NewDiff 0.90
JAWS_47_NewDiff 0.90
JAWS_17_NewDiff 0.91
JAWS_21_NewDiff 0.91
JAWS_37_NewDiff 0.91
JAWS_38_NewDiff 0.91
JAWS_48_NewDiff 0.91
JAWS_24_NewDiff 0.92
JAWS_29_NewDiff 0.92
JAWS_33_NewDiff 0.92
JAWS_34_NewDiff 0.92
JAWS_40_NewDiff 0.92
JAWS_4_NewDiff 0.93
JAWS_28_NewDiff 0.93
JAWS_39_NewDiff 0.93
JAWS_41_NewDiff 0.93
JAWS_3_NewDiff 0.94
JAWS_6_NewDiff 0.94
JAWS_7_NewDiff 0.94
JAWS_8_NewDiff 0.94
JAWS_11_NewDiff 0.94
JAWS_12_NewDiff 0.94
JAWS_13_NewDiff 0.94
JAWS_14_NewDiff 0.94
JAWS_15_NewDiff 0.94
JAWS_19_NewDiff 0.94
JAWS_23_NewDiff 0.94
JAWS_26_NewDiff 0.94
JAWS_44_NewDiff 0.94
JAWS_1_NewDiff 0.95
JAWS_5_NewDiff 0.95
JAWS_9_NewDiff 0.95
JAWS_10_NewDiff 0.95
JAWS_20_NewDiff 0.95
JAWS_27_NewDiff 0.95
JAWS_31_NewDiff 0.95
JAWS_36_NewDiff 0.95
JAWS_43_NewDiff 0.95
JAWS_16_NewDiff 0.96
JAWS_25_NewDiff 0.96
JAWS_35_NewDiff 0.96
JAWS_42_NewDiff 0.96
JAWS_2_NewDiff 0.97

Step 5: Determining the Number of Factors

Parallel analysis suggests that the number of factors = 10 and the number of components = NA Parallel analysis suggested 10 factors; Kaiser’s eigenvalue-over-1 rule suggested 3 factors.

Table 6. Factor-number comparison (all 48 items, 6 to 11 factors). Lower BIC = better; Any_h2_over_95 flags near-boundary communalities.
Factors RMSEA TLI BIC Max_communality Any_h2_over_95
6 0.060 0.815 -2810.6 0.780 FALSE
7 0.056 0.840 -2901.9 0.777 FALSE
8 0.053 0.856 -2893.3 0.995 TRUE
9 0.051 0.866 -2827.8 0.995 TRUE
10 0.049 0.877 -2762.5 0.875 FALSE
11 0.047 0.886 -2681.2 0.962 TRUE

Step 5b: Loadings structure for candidate solutions

Table 7a. 7-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 8
  1. Emotional content (+0.68); 26. Impact on lives (+0.60); 47. Graphic AV evidence (+0.58); 19. Emotions of parties (+0.53); 34. Media commentary (+0.50); 44. Sentencing (+0.49); 17. Self-rep litigants (+0.43); 25. Privilege/status (+0.42)
F2 7
  1. Camaraderie (+0.65); 29. Judicial leadership (+0.65); 41. Valued by leadership (+0.64); 21. Working w colleagues (+0.59); 42. Prof development (+0.49); 23. Leadership role (+0.47); 9. Distribution of work (+0.44)
F3 4
  1. Financial benefits (+0.63); 31. Leave entitlements (+0.58); 37. Security in court (+0.56); 38. Security beyond court (+0.50)
F4 3
  1. Working w court staff (+0.87); 24. Dealing w other staff (+0.81); 27. Judiciary & admin (+0.61)
F5 3
  1. Complex decisions (+0.69); 14. Assessing legal argument (+0.63); 12. Writing judgments (+0.54)
F6 3
  1. Workload (+0.76); 8. Pace of work (+0.75); 5. Control over day (+0.43)
F7 2
  1. Going on circuit (+0.86); 46. Living away (+0.79)
Table 7b. 8-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 8
  1. Emotional content (+0.68); 26. Impact on lives (+0.60); 47. Graphic AV evidence (+0.54); 19. Emotions of parties (+0.53); 44. Sentencing (+0.48); 34. Media commentary (+0.48); 17. Self-rep litigants (+0.43); 25. Privilege/status (+0.41)
F2 3
  1. Working w court staff (+0.83); 24. Dealing w other staff (+0.83); 27. Judiciary & admin (+0.64)
F3 4
  1. Financial benefits (+0.57); 37. Security in court (+0.54); 31. Leave entitlements (+0.51); 38. Security beyond court (+0.48)
F4 3
  1. Complex decisions (+0.73); 12. Writing judgments (+0.69); 14. Assessing legal argument (+0.65)
F5 3
  1. Working w colleagues (+0.83); 48. Camaraderie (+0.72); 10. Colleagues pace (+0.40)
F6 4
  1. Pace of work (+0.77); 4. Workload (+0.76); 5. Control over day (+0.45); 6. Adequacy of breaks (+0.43)
F7 2
  1. Judicial leadership (+0.75); 41. Valued by leadership (+0.70)
F8 2
  1. Living away (+1.01); 45. Going on circuit (+0.66)
Table 7c. 9-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 8
  1. Emotional content (+0.65); 26. Impact on lives (+0.57); 47. Graphic AV evidence (+0.55); 19. Emotions of parties (+0.49); 34. Media commentary (+0.47); 44. Sentencing (+0.46); 17. Self-rep litigants (+0.40); 25. Privilege/status (+0.40)
F2 3
  1. Complex decisions (+0.74); 12. Writing judgments (+0.69); 14. Assessing legal argument (+0.65)
F3 3
  1. Dealing w other staff (+0.83); 22. Working w court staff (+0.81); 27. Judiciary & admin (+0.64)
F4 2
  1. Working w colleagues (+0.81); 48. Camaraderie (+0.72)
F5 4
  1. Workload (+0.76); 8. Pace of work (+0.74); 5. Control over day (+0.42); 6. Adequacy of breaks (+0.42)
F6 2
  1. Judicial leadership (+0.76); 41. Valued by leadership (+0.69)
F7 3
  1. Financial benefits (+0.62); 31. Leave entitlements (+0.45); 32. Salary (+0.41)
F8 2
  1. Security in court (+0.84); 38. Security beyond court (+0.67)
F9 2
  1. Living away (+1.01); 45. Going on circuit (+0.66)

Step 5c: Notes on each factor

Note to self. Communality is the proportion of an item’s variance that the factor explains and the boundary (.995) is like 1 and means that the factors all for all of an item’s variance. Should be impossible because of measurement error and noise.

The 7-factor solution (Table 7a) has the best BIC but merges conceptually distinct content. First, its second factor combined collegial relationships (items 21 and 48) with leadership and recognition (items 29 and 41). Second, it combined physical safety (items 37 and 38) with financial and employment conditions (items 33 and 31). The 7-factor solution is therefore theoretically weak despite having the best BIC

The 8-factor solution (Table 7b) resolved the collegial and leadership merge at 7 factors. It separates collegial relationships (items 21 and 48) from leadership and recognition (items 29 and 41). However, it retained the less interpretable merge of physical safety (items 37 and 38) with financial and employment conditions (items 33 and 31) in a single factor. It also showed a near-boundary communality, with item 46 (living away) loading above 1.0 on the two-item travel factor, indicating that factor is so mewhat weakly determined.

Note. The 9-factor solution (Table 7c) resolves both merges in the other lower solutions: collegial relationships (items 21 and 48) becomes seperated from leadership and recognition (items 29 and 41), and physical safety (items 37 and 38) from financial and employment conditions (items 32, 33 and 31). I feel ike the seperation here gives it the strongest theoretical interpretion. Although the cost is parsimony: four factors are defined by only two items each, and the two-item travel factor (items 45 and 46) is still showing a near-boundary loading (item 46 above 1.0).

Step 6: Solution Diagnostics (9 factors)

Highest communality = 0.995 ( JAWS_46_NewDiff )

RMSEA = 0.051 (90% CI 0.048 to 0.054 ) | TLI = 0.866 | BIC = -2827.8

Table 8. Communalities (h2) by item, 9-factor solution, sorted high to low
Item Communality
  1. Living away
0.995
  1. Workload
0.791
  1. Working w court staff
0.760
  1. Valued by leadership
0.754
  1. Working w colleagues
0.742
  1. Dealing w other staff
0.731
  1. Pace of work
0.723
  1. Security in court
0.722
  1. Judicial leadership
0.704
  1. Complex decisions
0.700
  1. Camaraderie
0.664
  1. Writing judgments
0.617
  1. Control over day
0.582
  1. Emotional content
0.581
  1. Assessing legal argument
0.574
  1. Impact on lives
0.570
  1. Security beyond court
0.567
  1. Financial benefits
0.558
  1. Leave entitlements
0.552
  1. Adequacy of breaks
0.551
  1. Judiciary & admin
0.535
  1. Adequacy of staffing
0.532
  1. Prof development
0.521
  1. Contribution to society
0.505
  1. Going on circuit
0.495
  1. Distribution of work
0.493
  1. Emotions of parties
0.469
  1. Sentencing
0.452
  1. Work consistent w values
0.447
  1. Leadership role
0.408
  1. Physical environment
0.395
  1. Isolation of role
0.385
  1. Assessing witnesses
0.371
  1. Media commentary
0.364
  1. Quality of legal rep
0.363
  1. Graphic AV evidence
0.356
  1. Colleagues pace
0.348
  1. Interruptions
0.340
  1. Judicial education
0.340
  1. Privilege/status
0.314
  1. Support from A-G
0.312
  1. IT resources
0.286
  1. Salary
0.281
  1. Appellate review
0.267
  1. Activities outside court
0.265
  1. Self-rep litigants
0.263
  1. Rehab programs
0.233
  1. Managing juries
0.186

Note. Fit was acceptable for a long multi-factor instrument (RMSEA = .051, TLI = .866). The solution is admissible (no communality reached 1, so no Heywood case). Item 46 (living away) had the highest communality (.995). This reflected the weakly determined two-item travel factor rather than a problem with the item (i.e., almost all the variance loaded onto 46?). Another potential problem is Item 18 (managing juries) had the lowest (.19), which is consistent with its low response rate and weak correlations noted earlier.

Step 7: Factor Loadings and Item Classification

Items were assigned to a factor where the primary loading was >= .40 and no secondary loading reached .30. Cross-loaders (secondary >= .30) and orphans (no loading >= .30) are reported but excluded from factor membership.

Table 9. Item classification (primary >= .40, secondary < .30)
Item Factor Primary Secondary Status
  1. Living away
Circuit & Travel 1.01 0.03 Assigned
  1. Going on circuit
Circuit & Travel 0.66 0.08 Assigned
  1. Complex decisions
Cognitive Demands 0.74 0.11 Assigned
  1. Writing judgments
Cognitive Demands 0.69 0.16 Assigned
  1. Assessing legal argument
Cognitive Demands 0.65 0.12 Assigned
  1. Assessing witnesses
Cognitive Demands 0.40 0.25 Below .40 threshold
  1. Appellate review
Cognitive Demands 0.27 0.24 Orphan (no salient loading)
  1. Working w colleagues
Collegial Community 0.81 0.07 Assigned
  1. Camaraderie
Collegial Community 0.72 0.12 Assigned
  1. Colleagues pace
Collegial Community 0.39 0.17 Below .40 threshold
  1. Leadership role
Collegial Community 0.33 0.21 Below .40 threshold
  1. Judicial education
Collegial Community 0.33 0.18 Below .40 threshold
  1. Activities outside court
Collegial Community 0.28 0.21 Orphan (no salient loading)
  1. Work consistent w values
Collegial Community 0.28 0.21 Orphan (no salient loading)
  1. Dealing w other staff
Court Staff Relationships 0.83 0.05 Assigned
  1. Working w court staff
Court Staff Relationships 0.81 0.13 Assigned
  1. Judiciary & admin
Court Staff Relationships 0.64 0.17 Assigned
  1. Adequacy of staffing
Court Staff Relationships 0.33 0.29 Below .40 threshold
  1. IT resources
Court Staff Relationships 0.23 0.18 Orphan (no salient loading)
  1. Emotional content
Emotional Weight & Exposure 0.65 0.16 Assigned
  1. Impact on lives
Emotional Weight & Exposure 0.57 0.22 Assigned
  1. Graphic AV evidence
Emotional Weight & Exposure 0.55 0.13 Assigned
  1. Emotions of parties
Emotional Weight & Exposure 0.49 0.15 Assigned
  1. Media commentary
Emotional Weight & Exposure 0.47 0.12 Assigned
  1. Sentencing
Emotional Weight & Exposure 0.46 0.26 Assigned
  1. Self-rep litigants
Emotional Weight & Exposure 0.40 0.10 Assigned
  1. Privilege/status
Emotional Weight & Exposure 0.40 0.18 Assigned
  1. Rehab programs
Emotional Weight & Exposure 0.36 0.15 Below .40 threshold
  1. Contribution to society
Emotional Weight & Exposure 0.33 0.30 Cross-loader (excluded)
  1. Isolation of role
Emotional Weight & Exposure 0.27 0.15 Orphan (no salient loading)
  1. Support from A-G
Emotional Weight & Exposure 0.23 0.22 Orphan (no salient loading)
  1. Financial benefits
Employment Perks 0.62 0.12 Assigned
  1. Leave entitlements
Employment Perks 0.45 0.19 Assigned
  1. Salary
Employment Perks 0.41 0.14 Assigned
  1. Quality of legal rep
Employment Perks 0.31 0.24 Below .40 threshold
  1. Physical environment
Employment Perks 0.26 0.19 Orphan (no salient loading)
  1. Judicial leadership
Leadership & Recognition 0.76 0.07 Assigned
  1. Valued by leadership
Leadership & Recognition 0.69 0.09 Assigned
  1. Distribution of work
Leadership & Recognition 0.31 0.30 Cross-loader (excluded)
  1. Prof development
Leadership & Recognition 0.30 0.22 Orphan (no salient loading)
  1. Managing juries
Leadership & Recognition 0.19 0.18 Orphan (no salient loading)
  1. Security in court
Safety & Security 0.84 0.05 Assigned
  1. Security beyond court
Safety & Security 0.67 0.08 Assigned
  1. Workload
Workload & Pace 0.76 0.13 Assigned
  1. Pace of work
Workload & Pace 0.74 0.10 Assigned
  1. Control over day
Workload & Pace 0.42 0.31 Cross-loader (excluded)
  1. Adequacy of breaks
Workload & Pace 0.42 0.26 Assigned
  1. Interruptions
Workload & Pace 0.31 0.24 Below .40 threshold

Not assigned:

  • Orphan variabless (no loading >= .30): 2. Physical environment; 3. IT resources; 18. Managing juries; 20. Isolation of role; 28. Appellate review; 30. Activities outside court; 35. Work consistent w values; 39. Support from A-G; 42. Prof development
  • Cross-loaders (secondary >= .30): 5. Control over day; 9. Distribution of work; 36. Contribution to society
  • Primary .30 to .39: 1. Adequacy of staffing; 7. Interruptions; 10. Colleagues pace; 13. Assessing witnesses; 16. Quality of legal rep; 23. Leadership role; 40. Rehab programs; 43. Judicial education
Table 10. Factor summary (assigned items only, primary >= .40, secondary < .30)
Factor N Items Items % Variance
Emotional Weight & Exposure 8
  1. Emotional content; 17. Self-rep litigants; 19. Emotions of parties; 25. Privilege/status; 26. Impact on lives; 34. Media commentary; 44. Sentencing; 47. Graphic AV evidence
16.3%
Cognitive Demands 3
  1. Complex decisions; 12. Writing judgments; 14. Assessing legal argument
12.3%
Court Staff Relationships 3
  1. Working w court staff; 24. Dealing w other staff; 27. Judiciary & admin
12.3%
Collegial Community 2
  1. Working w colleagues; 48. Camaraderie
11.6%
Workload & Pace 3
  1. Workload; 6. Adequacy of breaks; 8. Pace of work
11.3%
Leadership & Recognition 2
  1. Judicial leadership; 41. Valued by leadership
9.8%
Employment Perks 3
  1. Leave entitlements; 32. Salary; 33. Financial benefits
9.4%
Safety & Security 2
  1. Security in court; 38. Security beyond court
9.1%
Circuit & Travel 2
  1. Going on circuit; 46. Living away
7.8%

Step 8: Rotation Check

Table 11. Cross-loaders (secondary >= .30) under oblique vs orthogonal rotation
Rotation Cross_loaders
Oblimin (oblique) 3
Varimax (orthogonal) 15

Step 9: Factor Correlations

Table 12. Factor correlation matrix (Phi)
Emotional Weight & Exposure Cognitive Demands Court Staff Relationships Collegial Community Workload & Pace Leadership & Recognition Employment Perks Safety & Security Circuit & Travel
Emotional Weight & Exposure 1.00 0.46 0.36 0.23 0.39 0.19 0.19 0.30 0.29
Cognitive Demands 0.46 1.00 0.33 0.39 0.41 0.23 0.23 0.27 0.20
Court Staff Relationships 0.36 0.33 1.00 0.50 0.32 0.38 0.25 0.40 0.17
Collegial Community 0.23 0.39 0.50 1.00 0.21 0.46 0.29 0.34 0.18
Workload & Pace 0.39 0.41 0.32 0.21 1.00 0.35 0.35 0.34 0.34
Leadership & Recognition 0.19 0.23 0.38 0.46 0.35 1.00 0.41 0.37 0.22
Employment Perks 0.19 0.23 0.25 0.29 0.35 0.41 1.00 0.45 0.22
Safety & Security 0.30 0.27 0.40 0.34 0.34 0.37 0.45 1.00 0.29
Circuit & Travel 0.29 0.20 0.17 0.18 0.34 0.22 0.22 0.29 1.00

Of 36 factor correlations, 19 exceeded .30 (mean |r| = 0.31 , strongest = 0.5 ).

The factor correlations (see Table 12) show substantial interrelationships(19 of 36 exceeded .30; strongest .50, between Collegial Community and Court Staff Relationships), which confirms the factors are not independent (good justification for oblim rotation I used [Field, pg. 788]). Consistent with this, the orthogonal (varimax) solution produced far more cross-loaders than the oblique solution (15 vs 3; Table 11) because forcing uncorrelated factors smears items across them. Oblique (oblimin) rotation is justified.

Step 10: Residual Analysis

Figure 1. Residual analysis (9-factor solution)
Statistic Value
Total residuals 1128
Residuals with &#124;value&#124; > .05 95
Percentage > .05 8.4%
RMSR (root mean square residual) 0.03
Mean residual 0.0002

The 9-factor solution reproduced the observed correlations well. Of 1,128 residuals, only 8.4% exceeded .05 (i.e., how many pairs of variables have a difference greater than .05 between their real-world correlation and the model’s predicted correlation). Field said this is below the 50% that would indicate poor fit, and the root mean square residual (.03) was below the .05 level considered good. Residuals were centred on zero (mean = .0002) and symmetrically distributed (see Figure 1), so no systematic over- or under-prediction of the correlations.

Exploratory Factor Analysis: JAWS Most Satisfying

expand

Step 1: Sample Size

Total sample: 602

Sample retained (i.e., answered > 50% of the 48 satisfaction items): N = 576 ( 26 excluded for more than 50% missing).

Cases per item (final n ÷ n of items) = 12

Table 1. Ten satisfaction items with the lowest valid N
Item Valid_N
  1. Managing juries
238
  1. Living away
383
  1. Going on circuit
423
  1. Leadership role
459
  1. Activities outside court
483
  1. Rehab programs
487
  1. Colleagues pace
493
  1. Sentencing
502
  1. Graphic AV evidence
503
  1. Security beyond court
520

Step 2: Correlation Matrix

No item pairs correlated above .90 (Table 2).

Determinant = 0.000000002471427 (below Field’s threshold (> .00001), but expected with 48 items since the determinant shrinks with every correlated pair; no pairs above .90, so no items removed).

24.2 % of item correlations exceeded .30 (mean |r| = 0.24 ; strongest |r| = 0.66 ), indicating the items share sufficient common variance to factor analyse. This is lower than the NewDiff difference scores (43.4%), reflecting more weakly correlated raw satisfaction ratings.

Table 2. Ten strongest item-pair correlations
Item_1 Item_2 r
  1. Working w colleagues
  1. Camaraderie
0.66
  1. Working w court staff
  1. Dealing w other staff
0.65
  1. Judicial leadership
  1. Valued by leadership
0.65
  1. Workload
  1. Pace of work
0.64
  1. Security in court
  1. Security beyond court
0.59
  1. Emotional content
  1. Emotions of parties
0.56
  1. Going on circuit
  1. Living away
0.56
  1. Complex decisions
  1. Writing judgments
0.55
  1. Complex decisions
  1. Assessing legal argument
0.54
  1. Control over day
  1. Adequacy of breaks
0.52

Step 3: Bartlett’s Test of Sphericity

Table 4. Bartlett’s test of sphericity
Statistic Value
Chi-square 11,062.01
df 1128
p-value < .001

Bartlett’s test of sphericity was significant, chi-square(1128) = 11,062.01, p < .001 (Table 4), indicating the correlation matrix differs from an identity matrix (i.e., correlation is expected to be 0) and the items are sufficiently interrelated for factor analysis.

Step 4: Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy

Overall MSA = 0.91 | item MSAs ranged from 0.65 to 0.96

Table 5. Per-item KMO (measure of sampling adequacy), sorted lowest to highest
Item MSA
  1. Managing juries
0.65
  1. Going on circuit
0.69
  1. Living away
0.79
  1. Financial benefits
0.85
  1. Salary
0.86
  1. Graphic AV evidence
0.87
  1. Interruptions
0.88
  1. Self-rep litigants
0.88
  1. Sentencing
0.88
  1. Working w court staff
0.89
  1. Dealing w other staff
0.89
  1. Appellate review
0.89
  1. Security beyond court
0.89
  1. Camaraderie
0.89
  1. Emotions of parties
0.90
  1. Working w colleagues
0.90
  1. Security in court
0.90
  1. Valued by leadership
0.90
  1. Colleagues pace
0.91
  1. Writing judgments
0.91
  1. Activities outside court
0.91
  1. Media commentary
0.91
  1. Control over day
0.92
  1. Emotional content
0.92
  1. Leadership role
0.92
  1. Judicial leadership
0.92
  1. Leave entitlements
0.92
  1. Support from A-G
0.92
  1. Adequacy of staffing
0.93
  1. Workload
0.93
  1. Pace of work
0.93
  1. Assessing witnesses
0.93
  1. Privilege/status
0.93
  1. Impact on lives
0.93
  1. Judiciary & admin
0.93
  1. Contribution to society
0.93
  1. Rehab programs
0.93
  1. IT resources
0.94
  1. Adequacy of breaks
0.94
  1. Complex decisions
0.94
  1. Assessing legal argument
0.94
  1. Isolation of role
0.94
  1. Judicial education
0.94
  1. Distribution of work
0.95
  1. Quality of legal rep
0.95
  1. Physical environment
0.96
  1. Work consistent w values
0.96
  1. Prof development
0.96

  • Sampling adequacy for satisfaction is also verified with KOM, with an overall MSA of 0.91, well above the .50 minimum and in the ‘superb’ rang, so also suitable for factor analysis.

  • three lowest (items 18, 45, and 46) were the items with the smallest n of responses all.

Step 5: Determining the Number of Factors

Parallel analysis suggests that the number of factors = 10 and the number of components = NA Parallel analysis suggested 10 factors; Kaiser’s eigenvalue-over-1 rule suggested 4 factors.

Table 6. Factor-number comparison (all 48 items, 7 to 11 factors). Lower BIC = better; Any_h2_over_95 flags near-boundary communalities.
Factors RMSEA TLI BIC Max_communality Any_h2_over_95
7 0.051 0.825 -3111.4 0.749 FALSE
8 0.049 0.843 -3081.8 0.995 TRUE
9 0.046 0.861 -3033.2 0.995 TRUE
10 0.043 0.875 -2959.5 0.995 TRUE
11 0.041 0.888 -2869.6 0.995 TRUE

Step 5b: Loadings structure for candidate solutions

Table 7a. 7-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 5
  1. Emotional content (+0.60); 19. Emotions of parties (+0.58); 47. Graphic AV evidence (+0.50); 26. Impact on lives (+0.47); 17. Self-rep litigants (+0.42)
F2 4
  1. Judicial leadership (+0.64); 41. Valued by leadership (+0.60); 48. Camaraderie (+0.45); 42. Prof development (+0.41)
F3 4
  1. Security in court (+0.63); 38. Security beyond court (+0.57); 33. Financial benefits (+0.49); 31. Leave entitlements (+0.44)
F4 3
  1. Complex decisions (+0.64); 12. Writing judgments (+0.62); 14. Assessing legal argument (+0.61)
F5 3
  1. Working w court staff (+0.87); 24. Dealing w other staff (+0.70); 27. Judiciary & admin (+0.51)
F6 3
  1. Pace of work (+0.75); 4. Workload (+0.69); 5. Control over day (+0.45)
F7 3
  1. Going on circuit (+0.77); 46. Living away (+0.61); 18. Managing juries (+0.52)
Table 7b. 8-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 5
  1. Camaraderie (+0.65); 29. Judicial leadership (+0.52); 21. Working w colleagues (+0.51); 41. Valued by leadership (+0.50); 42. Prof development (+0.48)
F2 4
  1. Emotions of parties (+0.57); 15. Emotional content (+0.56); 26. Impact on lives (+0.52); 17. Self-rep litigants (+0.46)
F3 3
  1. Security in court (+0.61); 38. Security beyond court (+0.55); 33. Financial benefits (+0.41)
F4 3
  1. Pace of work (+0.77); 4. Workload (+0.73); 5. Control over day (+0.45)
F5 3
  1. Complex decisions (+0.69); 12. Writing judgments (+0.68); 14. Assessing legal argument (+0.60)
F6 3
  1. Working w court staff (+0.92); 24. Dealing w other staff (+0.66); 27. Judiciary & admin (+0.45)
F7 2
  1. Living away (+0.76); 45. Going on circuit (+0.66)
F8 1
  1. Managing juries (+0.97)
Table 7c. 9-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 5
  1. Emotions of parties (+0.57); 26. Impact on lives (+0.55); 15. Emotional content (+0.52); 17. Self-rep litigants (+0.47); 40. Rehab programs (+0.41)
F2 2
  1. Security beyond court (+0.56); 37. Security in court (+0.56)
F3 2
  1. Judicial leadership (+0.73); 41. Valued by leadership (+0.64)
F4 3
  1. Pace of work (+0.80); 4. Workload (+0.74); 5. Control over day (+0.45)
F5 3
  1. Complex decisions (+0.68); 12. Writing judgments (+0.67); 14. Assessing legal argument (+0.61)
F6 2
  1. Camaraderie (+0.78); 21. Working w colleagues (+0.69)
F7 3
  1. Working w court staff (+0.85); 24. Dealing w other staff (+0.65); 27. Judiciary & admin (+0.47)
F8 1
  1. Managing juries (+0.97)
F9 2
  1. Living away (+0.72); 45. Going on circuit (+0.66)

Note on item 18 (managing juries). Across several diagnostics, item 18 doesn’t look like it belong with the other items. It had the lowest response rate (n = 238, 41% of the sample), because a large number of judicial officers do not sit with juries. It correlated weakly with the other items and had the lowest sampling adequacy (MSA = .65). Most tellingly, in the 8- and 9-factor solutions it formed its own single-item factor (loading .97, with no other item), which is what produced the near-boundary communality (.995) flagged in those solutions. An item that the model isolates into its own factor rather than grouping with any other shares too little common variance to be meaningfully placed. Ask Richard if we should delete and try again with this one?).

Step 5 (item 18 removed): candidate solutions

Parallel analysis suggests that the number of factors = 10 and the number of components = NA On 47 items (item 18 removed), parallel analysis suggested 10 factors.

Table 6b. Factor-number comparison, item 18 removed (47 items, 7 to 11 factors).
Factors RMSEA TLI BIC Max_communality Any_h2_over_95
7 0.048 0.850 -3128.2 0.795 FALSE
8 0.045 0.866 -3072.7 0.770 FALSE
9 0.043 0.878 -2983.3 0.824 FALSE
10 0.041 0.891 -2891.4 0.798 FALSE
11 0.038 0.905 -2806.6 0.775 FALSE
Table 7a (item 18 removed). 7-factor solution: items loading >= .40
Factor N_items Items
F1 6
  1. Judicial leadership (+0.61); 48. Camaraderie (+0.59); 41. Valued by leadership (+0.58); 42. Prof development (+0.50); 21. Working w colleagues (+0.47); 23. Leadership role (+0.41)
F2 5
  1. Emotions of parties (+0.59); 15. Emotional content (+0.58); 26. Impact on lives (+0.52); 17. Self-rep litigants (+0.46); 47. Graphic AV evidence (+0.42)
F3 3
  1. Security in court (+0.69); 38. Security beyond court (+0.62); 33. Financial benefits (+0.40)
F4 3
  1. Complex decisions (+0.69); 12. Writing judgments (+0.63); 14. Assessing legal argument (+0.63)
F5 3
  1. Pace of work (+0.75); 4. Workload (+0.70); 5. Control over day (+0.45)
F6 3
  1. Working w court staff (+0.89); 24. Dealing w other staff (+0.66); 27. Judiciary & admin (+0.46)
F7 2
  1. Living away (+0.77); 45. Going on circuit (+0.71)
Table 7b (item 18 removed). 8-factor solution: items loading >= .40
Factor N_items Items
F1 5
  1. Emotions of parties (+0.58); 15. Emotional content (+0.55); 26. Impact on lives (+0.52); 17. Self-rep litigants (+0.45); 47. Graphic AV evidence (+0.40)
F2 2
  1. Security in court (+0.65); 38. Security beyond court (+0.61)
F3 2
  1. Judicial leadership (+0.77); 41. Valued by leadership (+0.67)
F4 3
  1. Pace of work (+0.80); 4. Workload (+0.73); 5. Control over day (+0.45)
F5 2
  1. Camaraderie (+0.74); 21. Working w colleagues (+0.73)
F6 3
  1. Complex decisions (+0.67); 14. Assessing legal argument (+0.64); 12. Writing judgments (+0.62)
F7 3
  1. Working w court staff (+0.80); 24. Dealing w other staff (+0.66); 27. Judiciary & admin (+0.50)
F8 2
  1. Living away (+0.75); 45. Going on circuit (+0.71)
Table 7c (item 18 removed). 9-factor solution: items loading >= .40
Factor N_items Items
F1 5
  1. Emotions of parties (+0.59); 15. Emotional content (+0.55); 26. Impact on lives (+0.54); 17. Self-rep litigants (+0.46); 47. Graphic AV evidence (+0.41)
F2 3
  1. Pace of work (+0.77); 4. Workload (+0.72); 5. Control over day (+0.45)
F3 2
  1. Judicial leadership (+0.76); 41. Valued by leadership (+0.65)
F4 2
  1. Camaraderie (+0.73); 21. Working w colleagues (+0.72)
F5 2
  1. Security in court (+0.73); 38. Security beyond court (+0.70)
F6 3
  1. Complex decisions (+0.66); 12. Writing judgments (+0.62); 14. Assessing legal argument (+0.62)
F7 3
  1. Working w court staff (+0.81); 24. Dealing w other staff (+0.65); 27. Judiciary & admin (+0.49)
F8 2
  1. Financial benefits (+0.92); 32. Salary (+0.42)
F9 2
  1. Living away (+0.72); 45. Going on circuit (+0.71)

Note. With item 18 removed, all solutions (7 to 11 factors) were admissible with no near-boundary communalities. The 7-factor solution had the best BIC but merged conceptually distinct content, like mixing leadership and recognition (items 29, 41) with collegial relationships (items 48, 21), and physical safety (items 37, 38) with financial conditions (item 33). The 9-factor solution separated them into more distinct, interpretable factors while retaining the stable emotional, workload, cognitive, court-staff, and travel factors. It was retained as the best balance of parsimony and interpretability, and closely parallels the NewDiff structure. As with NewDiff, several factors rest on only two items and should be interpreted with caution.

Step 6: Solution Diagnostics (9 factors, item 18 removed)

Highest communality = 0.824 ( 33. Financial benefits )

RMSEA = 0.043 (90% CI 0.04 to 0.046 ) | TLI = 0.878 | BIC = -2983.3

Table 8. Communalities (h2) by item, 9-factor satisfaction solution (item 18 removed), sorted high to low
Item Communality
  1. Financial benefits
0.824
  1. Working w court staff
0.775
  1. Working w colleagues
0.661
  1. Camaraderie
0.654
  1. Workload
0.643
  1. Valued by leadership
0.639
  1. Judicial leadership
0.638
  1. Pace of work
0.628
  1. Security in court
0.595
  1. Living away
0.578
  1. Complex decisions
0.567
  1. Dealing w other staff
0.560
  1. Security beyond court
0.529
  1. Assessing legal argument
0.523
  1. Going on circuit
0.519
  1. Prof development
0.509
  1. Writing judgments
0.507
  1. Judiciary & admin
0.500
  1. Emotions of parties
0.483
  1. Adequacy of breaks
0.478
  1. Control over day
0.466
  1. Emotional content
0.456
  1. Leave entitlements
0.450
  1. Adequacy of staffing
0.448
  1. Impact on lives
0.445
  1. Contribution to society
0.431
  1. Physical environment
0.413
  1. Work consistent w values
0.393
  1. Distribution of work
0.377
  1. Leadership role
0.359
  1. Sentencing
0.357
  1. Rehab programs
0.349
  1. IT resources
0.344
  1. Judicial education
0.343
  1. Quality of legal rep
0.321
  1. Assessing witnesses
0.316
  1. Salary
0.316
  1. Isolation of role
0.298
  1. Self-rep litigants
0.272
  1. Colleagues pace
0.265
  1. Graphic AV evidence
0.244
  1. Support from A-G
0.243
  1. Media commentary
0.241
  1. Privilege/status
0.238
  1. Activities outside court
0.230
  1. Appellate review
0.207
  1. Interruptions
0.205

Step 7: Factor Loadings and Item Classification

Items were assigned where the primary loading was >= .40 and no secondary loading reached .30. Cross-loaders and orphans are reported but excluded from membership.

Table 9. Item classification (primary >= .40, secondary < .30)
Item Factor Primary Secondary Status
  1. Living away
Circuit & Travel 0.72 0.09 Assigned
  1. Going on circuit
Circuit & Travel 0.71 0.04 Assigned
  1. Complex decisions
Cognitive Demands 0.66 0.08 Assigned
  1. Writing judgments
Cognitive Demands 0.62 0.12 Assigned
  1. Assessing legal argument
Cognitive Demands 0.62 0.10 Assigned
  1. Camaraderie
Collegial Community 0.73 0.09 Assigned
  1. Working w colleagues
Collegial Community 0.72 0.16 Assigned
  1. Leadership role
Collegial Community 0.30 0.21 Orphan (no salient loading)
  1. Judicial education
Collegial Community 0.29 0.27 Orphan (no salient loading)
  1. Work consistent w values
Collegial Community 0.27 0.27 Orphan (no salient loading)
  1. Activities outside court
Collegial Community 0.21 0.17 Orphan (no salient loading)
  1. Working w court staff
Court Staff Relationships 0.81 0.14 Assigned
  1. Dealing w other staff
Court Staff Relationships 0.65 0.08 Assigned
  1. Judiciary & admin
Court Staff Relationships 0.49 0.30 Assigned
  1. Emotions of parties
Emotional Content & Exposure 0.59 0.14 Assigned
  1. Emotional content
Emotional Content & Exposure 0.55 0.14 Assigned
  1. Impact on lives
Emotional Content & Exposure 0.54 0.17 Assigned
  1. Self-rep litigants
Emotional Content & Exposure 0.46 0.08 Assigned
  1. Graphic AV evidence
Emotional Content & Exposure 0.41 0.19 Assigned
  1. Rehab programs
Emotional Content & Exposure 0.37 0.25 Below .40 threshold
  1. Privilege/status
Emotional Content & Exposure 0.35 0.17 Below .40 threshold
  1. Media commentary
Emotional Content & Exposure 0.33 0.17 Below .40 threshold
  1. Contribution to society
Emotional Content & Exposure 0.33 0.26 Below .40 threshold
  1. Assessing witnesses
Emotional Content & Exposure 0.31 0.22 Below .40 threshold
  1. Isolation of role
Emotional Content & Exposure 0.29 0.25 Orphan (no salient loading)
  1. Sentencing
Emotional Content & Exposure 0.29 0.25 Orphan (no salient loading)
  1. Appellate review
Emotional Content & Exposure 0.20 0.18 Orphan (no salient loading)
  1. Financial benefits
Employment Perks 0.92 0.02 Assigned
  1. Salary
Employment Perks 0.42 0.19 Assigned
  1. Leave entitlements
Employment Perks 0.33 0.24 Below .40 threshold
  1. Quality of legal rep
Employment Perks 0.22 0.14 Orphan (no salient loading)
  1. Judicial leadership
Leadership & Recognition 0.76 0.06 Assigned
  1. Valued by leadership
Leadership & Recognition 0.65 0.13 Assigned
  1. Distribution of work
Leadership & Recognition 0.36 0.22 Below .40 threshold
  1. Prof development
Leadership & Recognition 0.36 0.24 Below .40 threshold
  1. Support from A-G
Leadership & Recognition 0.28 0.15 Orphan (no salient loading)
  1. Adequacy of staffing
Leadership & Recognition 0.24 0.22 Orphan (no salient loading)
  1. Colleagues pace
Leadership & Recognition 0.20 0.20 Orphan (no salient loading)
  1. Security in court
Safety & Security 0.73 0.09 Assigned
  1. Security beyond court
Safety & Security 0.70 0.06 Assigned
  1. Physical environment
Safety & Security 0.31 0.17 Below .40 threshold
  1. Interruptions
Safety & Security 0.24 0.19 Orphan (no salient loading)
  1. IT resources
Safety & Security 0.22 0.20 Orphan (no salient loading)
  1. Pace of work
Workload & Pace 0.77 0.08 Assigned
  1. Workload
Workload & Pace 0.72 0.08 Assigned
  1. Control over day
Workload & Pace 0.45 0.19 Assigned
  1. Adequacy of breaks
Workload & Pace 0.36 0.31 Cross-loader (excluded)

Not assigned:

  • Orphans (no loading >= .30): none
  • Cross-loaders (secondary >= .30): 6. Adequacy of breaks
  • Primary .30 to .39: 2. Physical environment; 9. Distribution of work; 13. Assessing witnesses; 25. Privilege/status; 31. Leave entitlements; 34. Media commentary; 36. Contribution to society; 40. Rehab programs; 42. Prof development
Table 10. Factor summary (assigned items, primary >= .40, secondary < .30)
Factor N Items Items % Variance
Emotional Content & Exposure 5
  1. Emotional content; 17. Self-rep litigants; 19. Emotions of parties; 26. Impact on lives; 47. Graphic AV evidence
15.8%
Workload & Pace 3
  1. Workload; 5. Control over day; 8. Pace of work
12.7%
Leadership & Recognition 2
  1. Judicial leadership; 41. Valued by leadership
12%
Collegial Community 2
  1. Working w colleagues; 48. Camaraderie
11.8%
Safety & Security 2
  1. Security in court; 38. Security beyond court
11.5%
Cognitive Demands 3
  1. Complex decisions; 12. Writing judgments; 14. Assessing legal argument
10.9%
Court Staff Relationships 3
  1. Working w court staff; 24. Dealing w other staff; 27. Judiciary & admin
9.8%
Employment Perks 2
  1. Salary; 33. Financial benefits
8.7%
Circuit & Travel 2
  1. Going on circuit; 46. Living away
6.7%

I noticed one factor dominated in the NewDiff version but the satisfaction factors accounted for more evenly distributed variance (15.8% down to 6.7%). I think this is meant to happen with the more weakly and diffusely correlated raw satisfaction items.

Step 8: Rotation Check

Table 11. Cross-loaders (secondary >= .30) under oblique vs orthogonal rotation
Rotation Cross_loaders
Oblimin (oblique) 1
Varimax (orthogonal) 8

The orthogonal (varimax) solution showed more cross-loaders than the oblique solution (8 vs 1; Table 11), again tells us the factors are not independent.

Step 9: Factor Correlations

Table 12. Factor correlation matrix (Phi)
Emotional Content & Exposure Workload & Pace Leadership & Recognition Collegial Community Safety & Security Cognitive Demands Court Staff Relationships Employment Perks Circuit & Travel
Emotional Content & Exposure 1.00 0.37 0.20 0.25 0.27 0.27 0.34 0.10 0.26
Workload & Pace 0.37 1.00 0.40 0.27 0.40 0.32 0.27 0.36 0.24
Leadership & Recognition 0.20 0.40 1.00 0.48 0.39 0.27 0.33 0.40 0.11
Collegial Community 0.25 0.27 0.48 1.00 0.28 0.39 0.42 0.28 0.13
Safety & Security 0.27 0.40 0.39 0.28 1.00 0.20 0.30 0.41 0.18
Cognitive Demands 0.27 0.32 0.27 0.39 0.20 1.00 0.27 0.28 0.10
Court Staff Relationships 0.34 0.27 0.33 0.42 0.30 0.27 1.00 0.17 0.16
Employment Perks 0.10 0.36 0.40 0.28 0.41 0.28 0.17 1.00 0.14
Circuit & Travel 0.26 0.24 0.11 0.13 0.18 0.10 0.16 0.14 1.00

Of 36 factor correlations, 14 exceeded .30 (mean |r| = 0.28 , strongest = 0.48 ).

Note. The factors were moderately intercorrelated (14 of 36 correlations exceeded .30; Table 12). The correlations were weaker than for the NewDiff scores (19 of 36 above .30). Probably meant to happen with the more loosely correlated raw satisfaction items.

Exploratory Factor Analysis: JAWS Most stressful

expand

Step 1: Sample Size

Total participants: 602

Participants retained (i.e., answered > 50% of the 48 stress items): N = 576 ( 26 excluded for more than 50% missing).

Cases per item (final n ÷ n of items) = 12

Table 1. Ten stress items with the lowest valid N
Item Valid_N
  1. Managing juries
238
  1. Living away
385
  1. Going on circuit
427
  1. Leadership role
462
  1. Activities outside court
483
  1. Rehab programs
488
  1. Sentencing
502
  1. Graphic AV evidence
517
  1. Colleagues pace
520
  1. Financial benefits
541

Step 2: Correlation Matrix

No item pairs correlated above .90 (Table 2).

Determinant = 0.00000000002319306 (below Field’s threshold, expected with 48 items; no pairs above .90, so no items removed on this basis).

47.8 % of item correlations exceeded .30 (mean |r| = 0.3 ; strongest |r| = 0.77 ).

Table 2. Ten strongest item-pair correlations
Item_1 Item_2 r
  1. Workload
  1. Pace of work
0.77
  1. Going on circuit
  1. Living away
0.72
  1. Judicial leadership
  1. Valued by leadership
0.68
  1. Working w court staff
  1. Dealing w other staff
0.65
  1. Security in court
  1. Security beyond court
0.64
  1. Complex decisions
  1. Writing judgments
0.63
  1. Control over day
  1. Adequacy of breaks
0.61
  1. Complex decisions
  1. Assessing legal argument
0.60
  1. Dealing w other staff
  1. Judiciary & admin
0.60
  1. Workload
  1. Control over day
0.58

Step 3: Bartlett’s Test of Sphericity

Table 4. Bartlett’s test of sphericity
Statistic Value
Chi-square 13,667.92
df 1128
p-value < .001

Step 4: Kaiser-Meyer-Olkin (KMO)

Overall MSA = 0.94 | item MSAs ranged from 0.77 to 0.97

Table 5. Per-item KMO, sorted lowest to highest
Item MSA
  1. Managing juries
0.77
  1. Living away
0.84
  1. Activities outside court
0.86
  1. Going on circuit
0.86
  1. Working w colleagues
0.90
  1. Leadership role
0.90
  1. Salary
0.91
  1. Self-rep litigants
0.92
  1. Working w court staff
0.92
  1. Security in court
0.92
  1. Support from A-G
0.92
  1. Pace of work
0.93
  1. Colleagues pace
0.93
  1. Writing judgments
0.93
  1. Judicial leadership
0.93
  1. Financial benefits
0.93
  1. Media commentary
0.93
  1. Security beyond court
0.93
  1. Graphic AV evidence
0.93
  1. Camaraderie
0.93
  1. Workload
0.94
  1. Distribution of work
0.94
  1. Dealing w other staff
0.94
  1. Rehab programs
0.94
  1. Judicial education
0.94
  1. Sentencing
0.94
  1. IT resources
0.95
  1. Adequacy of breaks
0.95
  1. Complex decisions
0.95
  1. Assessing legal argument
0.95
  1. Emotional content
0.95
  1. Impact on lives
0.95
  1. Appellate review
0.95
  1. Leave entitlements
0.95
  1. Adequacy of staffing
0.96
  1. Physical environment
0.96
  1. Interruptions
0.96
  1. Assessing witnesses
0.96
  1. Emotions of parties
0.96
  1. Isolation of role
0.96
  1. Judiciary & admin
0.96
  1. Contribution to society
0.96
  1. Valued by leadership
0.96
  1. Prof development
0.96
  1. Control over day
0.97
  1. Quality of legal rep
0.97
  1. Privilege/status
0.97
  1. Work consistent w values
0.97

Step 5: Determining the Number of Factors

Parallel analysis suggests that the number of factors = 8 and the number of components = NA Parallel analysis suggested 8 factors; Kaiser’s eigenvalue-over-1 rule suggested 3 factors.

Table 6. Factor-number comparison (all 48 items, 7 to 11 factors). Lower BIC = better; Any_h2_over_95 flags near-boundary communalities.
Factors RMSEA TLI BIC Max_communality Any_h2_over_95
7 0.051 0.863 -3131.1 0.995 TRUE
8 0.048 0.879 -3106.9 0.851 FALSE
9 0.046 0.891 -3042.3 0.856 FALSE
10 0.044 0.899 -2943.4 0.995 TRUE
11 0.041 0.910 -2861.7 0.995 TRUE

Step 5b: Loadings structure for candidate solutions

Table 7a. 7-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 6
  1. Graphic AV evidence (+0.68); 15. Emotional content (+0.62); 34. Media commentary (+0.53); 44. Sentencing (+0.48); 26. Impact on lives (+0.43); 19. Emotions of parties (+0.41)
F2 5
  1. Dealing w other staff (+0.82); 22. Working w court staff (+0.73); 27. Judiciary & admin (+0.69); 1. Adequacy of staffing (+0.43); 3. IT resources (+0.41)
F3 5
  1. Financial benefits (+0.71); 31. Leave entitlements (+0.58); 32. Salary (+0.54); 37. Security in court (+0.50); 38. Security beyond court (+0.42)
F4 7
  1. Working w colleagues (+0.73); 48. Camaraderie (+0.64); 29. Judicial leadership (+0.52); 9. Distribution of work (+0.49); 41. Valued by leadership (+0.48); 10. Colleagues pace (+0.44); 23. Leadership role (+0.41)
F5 4
  1. Complex decisions (+0.66); 14. Assessing legal argument (+0.63); 12. Writing judgments (+0.54); 13. Assessing witnesses (+0.43)
F6 2
  1. Workload (+0.77); 8. Pace of work (+0.75)
F7 2
  1. Living away (+1.03); 45. Going on circuit (+0.69)
Table 7b. 8-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 7
  1. Complex decisions (+0.67); 14. Assessing legal argument (+0.67); 12. Writing judgments (+0.57); 13. Assessing witnesses (+0.50); 36. Contribution to society (+0.42); 44. Sentencing (+0.42); 26. Impact on lives (+0.41)
F2 4
  1. Dealing w other staff (+0.78); 22. Working w court staff (+0.75); 27. Judiciary & admin (+0.66); 1. Adequacy of staffing (+0.41)
F3 4
  1. Graphic AV evidence (+0.65); 15. Emotional content (+0.58); 34. Media commentary (+0.51); 19. Emotions of parties (+0.41)
F4 5
  1. Financial benefits (+0.66); 31. Leave entitlements (+0.51); 32. Salary (+0.50); 37. Security in court (+0.47); 38. Security beyond court (+0.44)
F5 2
  1. Valued by leadership (+0.73); 29. Judicial leadership (+0.70)
F6 2
  1. Pace of work (+0.76); 4. Workload (+0.76)
F7 2
  1. Living away (+0.92); 45. Going on circuit (+0.81)
F8 2
  1. Working w colleagues (+0.89); 48. Camaraderie (+0.42)
Table 7c. 9-factor solution: items loading >= .40 by factor
Factor N_items Items
F1 5
  1. Complex decisions (+0.67); 14. Assessing legal argument (+0.66); 12. Writing judgments (+0.60); 13. Assessing witnesses (+0.46); 28. Appellate review (+0.40)
F2 3
  1. Dealing w other staff (+0.78); 22. Working w court staff (+0.74); 27. Judiciary & admin (+0.66)
F3 5
  1. Emotional content (+0.62); 47. Graphic AV evidence (+0.61); 44. Sentencing (+0.47); 26. Impact on lives (+0.42); 19. Emotions of parties (+0.40)
F4 2
  1. Valued by leadership (+0.73); 29. Judicial leadership (+0.70)
F5 2
  1. Workload (+0.73); 8. Pace of work (+0.73)
F6 2
  1. Living away (+0.95); 45. Going on circuit (+0.78)
F7 3
  1. Security beyond court (+0.59); 37. Security in court (+0.58); 39. Support from A-G (+0.41)
F8 3
  1. Financial benefits (+0.56); 31. Leave entitlements (+0.50); 32. Salary (+0.40)
F9 2
  1. Working w colleagues (+0.89); 48. Camaraderie (+0.43)

Step 6: Solution Diagnostics (9 factors)

Highest communality = 0.856 ( 21. Working w colleagues )

RMSEA = 0.046 (90% CI 0.043 to 0.049 ) | TLI = 0.891 | BIC = -3042.3

Table 8. Communalities (h2) by item, 9-factor stress solution, sorted high to low
Item Communality
  1. Working w colleagues
0.856
  1. Living away
0.845
  1. Workload
0.782
  1. Valued by leadership
0.736
  1. Pace of work
0.732
  1. Dealing w other staff
0.678
  1. Going on circuit
0.668
  1. Complex decisions
0.652
  1. Working w court staff
0.651
  1. Judicial leadership
0.634
  1. Emotional content
0.610
  1. Control over day
0.589
  1. Writing judgments
0.588
  1. Adequacy of staffing
0.571
  1. Security beyond court
0.563
  1. Financial benefits
0.557
  1. Judiciary & admin
0.543
  1. Leave entitlements
0.536
  1. Security in court
0.536
  1. Assessing legal argument
0.533
  1. Adequacy of breaks
0.527
  1. Impact on lives
0.525
  1. Emotions of parties
0.521
  1. Sentencing
0.517
  1. Distribution of work
0.508
  1. Camaraderie
0.504
  1. Graphic AV evidence
0.489
  1. Contribution to society
0.458
  1. Media commentary
0.447
  1. Prof development
0.444
  1. Interruptions
0.431
  1. Isolation of role
0.425
  1. Assessing witnesses
0.414
  1. Support from A-G
0.400
  1. Work consistent w values
0.386
  1. Appellate review
0.383
  1. Physical environment
0.375
  1. Colleagues pace
0.365
  1. Privilege/status
0.345
  1. Leadership role
0.338
  1. Self-rep litigants
0.329
  1. Salary
0.326
  1. Managing juries
0.321
  1. Judicial education
0.315
  1. Quality of legal rep
0.295
  1. IT resources
0.289
  1. Rehab programs
0.284
  1. Activities outside court
0.168

Step 7: Factor Loadings and Item Classification

Items were assigned where the primary loading was >= .40 and no secondary loading reached .30. Cross-loaders and orphans are reported but excluded from membership.

Table 9. Item classification (primary >= .40, secondary < .30)
Item Factor Primary Secondary Status
  1. Living away
Circuit & Travel 0.95 0.07 Assigned
  1. Going on circuit
Circuit & Travel 0.78 0.10 Assigned
  1. Isolation of role
Circuit & Travel 0.26 0.17 Orphan (no salient loading)
  1. Complex decisions
Cognitive Demands 0.67 0.18 Assigned
  1. Assessing legal argument
Cognitive Demands 0.66 0.09 Assigned
  1. Writing judgments
Cognitive Demands 0.60 0.32 Cross-loader (excluded)
  1. Assessing witnesses
Cognitive Demands 0.46 0.22 Assigned
  1. Appellate review
Cognitive Demands 0.40 0.22 Assigned
  1. Contribution to society
Cognitive Demands 0.35 0.27 Below .40 threshold
  1. Self-rep litigants
Cognitive Demands 0.26 0.25 Orphan (no salient loading)
  1. Quality of legal rep
Cognitive Demands 0.20 0.15 Orphan (no salient loading)
  1. Working w colleagues
Collegial Community 0.89 0.03 Assigned
  1. Camaraderie
Collegial Community 0.43 0.27 Assigned
  1. Colleagues pace
Collegial Community 0.35 0.24 Below .40 threshold
  1. Dealing w other staff
Court Staff Relationships 0.78 0.09 Assigned
  1. Working w court staff
Court Staff Relationships 0.74 0.20 Assigned
  1. Judiciary & admin
Court Staff Relationships 0.66 0.10 Assigned
  1. Adequacy of staffing
Court Staff Relationships 0.40 0.22 Below .40 threshold
  1. IT resources
Court Staff Relationships 0.38 0.16 Below .40 threshold
  1. Judicial education
Court Staff Relationships 0.23 0.19 Orphan (no salient loading)
  1. Emotional content
Emotional Content & Exposure 0.62 0.15 Assigned
  1. Graphic AV evidence
Emotional Content & Exposure 0.61 0.16 Assigned
  1. Sentencing
Emotional Content & Exposure 0.47 0.35 Cross-loader (excluded)
  1. Impact on lives
Emotional Content & Exposure 0.42 0.34 Cross-loader (excluded)
  1. Emotions of parties
Emotional Content & Exposure 0.40 0.16 Assigned
  1. Rehab programs
Emotional Content & Exposure 0.33 0.17 Below .40 threshold
  1. Privilege/status
Emotional Content & Exposure 0.28 0.25 Orphan (no salient loading)
  1. Work consistent w values
Emotional Content & Exposure 0.21 0.20 Orphan (no salient loading)
  1. Activities outside court
Emotional Content & Exposure 0.17 0.15 Orphan (no salient loading)
  1. Financial benefits
Employment Perks 0.56 0.21 Assigned
  1. Leave entitlements
Employment Perks 0.50 0.17 Assigned
  1. Salary
Employment Perks 0.40 0.17 Assigned
  1. Managing juries
Employment Perks 0.30 0.25 Below .40 threshold
  1. Physical environment
Employment Perks 0.29 0.28 Orphan (no salient loading)
  1. Valued by leadership
Leadership & Recognition 0.73 0.08 Assigned
  1. Judicial leadership
Leadership & Recognition 0.70 0.10 Assigned
  1. Prof development
Leadership & Recognition 0.33 0.22 Below .40 threshold
  1. Distribution of work
Leadership & Recognition 0.32 0.31 Cross-loader (excluded)
  1. Leadership role
Leadership & Recognition 0.32 0.22 Below .40 threshold
  1. Security beyond court
Safety & Security 0.59 0.16 Assigned
  1. Security in court
Safety & Security 0.58 0.15 Assigned
  1. Support from A-G
Safety & Security 0.41 0.16 Assigned
  1. Media commentary
Safety & Security 0.38 0.37 Cross-loader (excluded)
  1. Workload
Workload & Pace 0.73 0.13 Assigned
  1. Pace of work
Workload & Pace 0.73 0.10 Assigned
  1. Control over day
Workload & Pace 0.34 0.32 Cross-loader (excluded)
  1. Adequacy of breaks
Workload & Pace 0.32 0.31 Cross-loader (excluded)
  1. Interruptions
Workload & Pace 0.32 0.24 Below .40 threshold

Not assigned:

  • Orphans (no loading >= .30): 2. Physical environment; 16. Quality of legal rep; 17. Self-rep litigants; 20. Isolation of role; 25. Privilege/status; 30. Activities outside court; 35. Work consistent w values; 43. Judicial education
  • Cross-loaders (secondary >= .30): 5. Control over day; 6. Adequacy of breaks; 9. Distribution of work; 12. Writing judgments; 26. Impact on lives; 34. Media commentary; 44. Sentencing
  • Primary .30 to .39: 1. Adequacy of staffing; 3. IT resources; 7. Interruptions; 10. Colleagues pace; 18. Managing juries; 23. Leadership role; 36. Contribution to society; 40. Rehab programs; 42. Prof development
Table 10. Factor summary (assigned items, primary >= .40, secondary < .30)
Factor N Items Items % Variance
Cognitive Demands 4
  1. Complex decisions; 13. Assessing witnesses; 14. Assessing legal argument; 28. Appellate review
14.7%
Court Staff Relationships 3
  1. Working w court staff; 24. Dealing w other staff; 27. Judiciary & admin
13.9%
Emotional Content & Exposure 3
  1. Emotional content; 19. Emotions of parties; 47. Graphic AV evidence
12.8%
Leadership & Recognition 2
  1. Judicial leadership; 41. Valued by leadership
11.6%
Workload & Pace 2
  1. Workload; 8. Pace of work
10.7%
Circuit & Travel 2
  1. Going on circuit; 46. Living away
10%
Safety & Security 3
  1. Security in court; 38. Security beyond court; 39. Support from A-G
9.2%
Employment Perks 3
  1. Leave entitlements; 32. Salary; 33. Financial benefits
8.9%
Collegial Community 2
  1. Working w colleagues; 48. Camaraderie
8.2%

Step 8: Rotation Check

Table 11. Cross-loaders (secondary >= .30) under oblique vs orthogonal rotation
Rotation Cross_loaders
Oblimin (oblique) 7
Varimax (orthogonal) 20

Step 9: Factor Correlations

Table 12. Factor correlation matrix (Phi)
Cognitive Demands Court Staff Relationships Emotional Content & Exposure Leadership & Recognition Workload & Pace Circuit & Travel Safety & Security Employment Perks Collegial Community
Cognitive Demands 1.00 0.34 0.46 0.21 0.41 0.34 0.20 0.14 0.22
Court Staff Relationships 0.34 1.00 0.31 0.49 0.33 0.26 0.33 0.28 0.42
Emotional Content & Exposure 0.46 0.31 1.00 0.26 0.31 0.41 0.29 0.11 0.13
Leadership & Recognition 0.21 0.49 0.26 1.00 0.36 0.33 0.38 0.38 0.43
Workload & Pace 0.41 0.33 0.31 0.36 1.00 0.35 0.29 0.29 0.21
Circuit & Travel 0.34 0.26 0.41 0.33 0.35 1.00 0.37 0.25 0.16
Safety & Security 0.20 0.33 0.29 0.38 0.29 0.37 1.00 0.35 0.27
Employment Perks 0.14 0.28 0.11 0.38 0.29 0.25 0.35 1.00 0.23
Collegial Community 0.22 0.42 0.13 0.43 0.21 0.16 0.27 0.23 1.00

Of 36 factor correlations, 19 exceeded .30 (mean |r| = 0.3 , strongest = 0.49 ).

Note. The factors were moderately intercorrelated (19 of 36 correlations exceeded .30; mean |r| = .30, strongest = .49; Table 12). The stress factors were intercorrelated to a similar degree as the NewDiff factors (19 of 36 above .30, mean |r| = .31) and more so than the satisfaction factors (14 of 36, mean |r| = .28).

JAWS Difference by Distress Heat Map: Gender

expand

JAWS Difference by Distress Heat Map: Wellbeing Measures

expand

Factor Analysis: JAWS Difference by Distress (Excluding Lower Court Judges)

expand

N (Supreme + District, after >50% missing filter): 264 | cases per item = 5.5

Highest communality = 0.995 ( JAWS_45_NewDiff ) RMSEA = 0.058 | TLI = 0.82 | BIC = -2700.2

Higher-courts N = 264

Item 18 valid N = 206 of 264 ( 78 %)

Item 18 (managing juries) loadings across the 9 factors:

ML9 ML1 ML8 ML2 ML5 ML7 ML3 ML6 ML4 0.06 0.15 0.13 -0.19 -0.10 0.18 0.14 0.26 0.11

Highest loading = 0.26 | second = 0.19 Communality (h2) = 0.275

Item 18 classification: Orphan variable (no salient loading)