development and validation of an improved hardiness tool ...iamps.org/papers/2019_bartone.pdf ·...
TRANSCRIPT
Development and Validation of an Improved Hardiness Tool:
The Hardiness Resilience Gauge
COL (Ret.) Paul T. Bartone, Ph.D.
Institute for National Security Studies
National Defense University
Washington, DC, USA
Kelly McDonald, Ph.D., Jonathan Stermac, Ph.D., Manalo Escobar, Ph.D.,
Justin McNeil, Ph.D. & Steven Stein, Ph.D.
MHS – Multi Health Systems
Toronto, CANADA
IAMPS, 7-10 May 2019 Sarajevo Contact: [email protected]
1992-1993 UNPROFOR to the former Yugoslavia
U.S. Army field hospital – MASH at Zagreb airport
Personality Hardiness
A general tendency to view the world as interesting &
meaningful; to see self as able to exert control; to see change as
opportunity to learn and grow
Commitment (vs. alienation)
… ability to feel deeply involved in life – including
work, social world, physical world, and oneself
Control (vs. powerlessness)
…belief you can control or influence events and outcomes
Challenge (vs. threat)
…sees change as an exciting challenge & chance to learn
5
Measuring hardiness
•1979 Kobasa: 18 scales from 5 different tests; 101 items,
all negatively keyed, poor factor structure
•1984 Bartone dissertation: 76 50 items (20 CM, 20 CO, 10 CH), all
negatively keyed)
•1985 Bartone post doc: 50 items (16 CM, 17 CO, 17 CH,
39 neg.) becomes the Personal Views Survey PVS used by
Maddi - Hardiness Institute
•1989 Bartone JNMD: 45 items (15 CM, 15 CO, 15 CH,
30 neg. becomes the Dispositional Resilience Scale DRS
6
•1991 Bartone, APS: 30-item short DRS, 15 each CM, CO, &
CH, fully balanced for pos and neg
•1995 Bartone, APS: 15-item short DRS, 5 each CM, CO, CH
with 4 neg. keyed
•1998 Bartone, West Point: DRS-15 v.2, minor refinements and
5 neg. keyed
•2007 Bartone, Fulbright: DRS-15 v.3, improved CO scale, 6
neg. keyed (Bartone, 2013; Hystad et al., 2010)
•2017 DRS assigned to MHS – Multi Health Systems
Measuring hardiness
7
2017-18 MHS revision
Goal: Building on the DRS-15, create a more comprehensive measure of hardiness with improved construct validity and reliability
15 DRS original items21 new items written (Bartone, McDonald, MHS)
8 Challenge6 Control7 Commitment
____36 items total subjected to further analysis
8
• Normative sample collected June 2018 (TurkPrime)
• Sample total (N = 2016); N=1500 after screening out spurious responses (143) and census matching
• U.S. census-matched within 1 % on:
• Race/Ethnicity
• Geographic Region
• Education
• Employment Status
• Gender
• Age
9
10
Table 4.1. Demographic Characteristics of the Normative Sample
Demographic Frequency N %
Gender Male 750 50.0
Female 750 50.0
Age Group (Years)
18–24 300 20.0
25–34 300 20.0 35–54 300 20.0
55–64 300 20.0
65+ 300 20.0 Median Age (SD) 43 (18.7)
Racial/Ethnic Group
Asian 82 5.5
Black 177 11.8 Hispanic 241 16.0
White 957 63.8 Other 43 2.9
Geographic Region
Northeast U.S. 264 17.6
Midwest U.S. 372 24.8
South U.S. 528 35.2
West U.S.
336 22.4
Education Level
No High School Diploma 147 9.8
High School Graduate 350 23.3
Some College or Associate Degree 505 33.7
Bachelor’s Degree 309 20.6
Graduate School 189 12.6
Employment Status
Employed or Self-Employed: Full-Time
589 39.3
Employed or Self-Employed: Part-Time
190 12.7
Retired 382 25.4
Student: Full-Time 167 11.1 Student: Part-Time 43 2.9
Unemployed 129 8.6
Total 1500
Items selected based on:
• Item Response distributions • Item-total correlations• Item loadings on factors (CFA)• IRT (Item Response Theory) item curves• Confirmatory Factor Analysis (CFA) Structure• Scale Reliabilities
11
Item Response Theory: A “Good” Item
Ability (hardiness) Level (Theta)
Responses (0-3)
Proportion of People Who Selected Responses at Each Level of Hardiness
Ideally, we want no overlapbetween curves (items should discriminate based on ability)
We want curves, not straight lines (people with the ability should always answer the same way)
12
Item Response Theory Trace Lines
13
Item Response Trace Lines: Challenge
14
Confirmatory Factor AnalysisFit Indices
• Chi-square/p-value
– Crude test of fit; highly sensitive to sample size
• Fit indices: CFI / TLI
– Higher is better
– Want values above .90
• RMSEA, SRMR
– Lower is better
– Want values below .1
15
1. Single Hardiness Factor
Model Chi-Square df P CFI TLI RMSEA SRMR
1 19063.37 594 <.001 .926 .922 .124 .107
2 123190.63 594 <.001 .510 .481 .320 .271
2. Three Uncorrelated Factors (Orthogonal)
16
CFA with 36 items
Model Chi-Square
df P CFI TLI RMSEA SRMR
3 14815.77 594 <.001 .943 .939 .109 .097
3. Hierarchical Model
17
CFA with 36 items
28 Items selected based on all criteria
• 10 Challenge items “Changes in routine are interesting to me”
“Mistakes are opportunities to learn and improve”
• 8 Control items“I am responsible for my own success in life”
“I am confident I can accomplish whatever I set out to do”
• 10 Commitment items “I look forward to my daily activities”
“I have a clear sense of purpose in my life”18
Model Chi-Square
df P CFI TLI RMSEA SRMR
3b 3351.03 347 <.001 .98 .98 .08 .06
3b. Final Hierarchical Model
19
CFA with final 28 items
Model Param-eters
Chi-Square
df P CFI TLI RMSEA SRMR
1 112 6045.7 350 <.001 .96 .96 .10 .08
2 112 76949.4 350 <.001 .53 .49 .38 .31
3 115 3351.03 347 <.001 .98 .98 .08 .06
Final fit statistics on three different models, 28 HRG items
Model 1: A single general hardiness factorModel 2: Three uncorrelated factorsModel 3: Hierarchical- 3 correlated factors under a general factor
20
Cronbach’sAlpha
(N=1500)
3 week Test-Retest(N=168)
Total Hardiness .93 .81
Challenge .85 .80
Control .84 .74
Commitment .89 .79
HRG Scale Reliability
21
HRGTask
focusedcoping1
Emotionfocused coping1
SWLS2 Thrive in challenging situations3
MaslachEfficacy4
Maslach Cynicism4
Hardiness .68 -30 .48 .57 .44 -.37
Challenge .62 -.26 .33 .58 .35 -.27*
Control .55 -.22 .39 .44 .41 -.27*
Commitment .61 -.31 .56 .50 .43 -.47
HRG Convergent and Discriminant Validity
All correlations significant at p<.001 unless otherwise indicated; *p<.011 N=1461; CISS - Coping with Stressful Situations Inventory, Endler & Parker, 19902 N=1461; SWLS – Satisfaction with Life Scale, Diener et al., 1985 3 N=1461; Single item 7-point likert scale “I believe I can thrive in challenging situations”4 N=113 employed adults; Maslach Burnout Inventory; Maslach, Jackson & Leiter, 1996
22
For more info: www.mhs.com/hrg
23
BACKUP
24
Demographics – Group Differences
25
• Sex – no differences
• Age – no differences 25-34 age group ^ than 65+ in challenge (h2=.004)
• Small ethnic/race differencesCompared to White, Hispanic & Black slightly higher in hardiness (Cohen’s d -0.12, .-0.38)
26
Table 4.12. Differences between Racial/Ethnic Groups
White
H
(N = 241)
Hispanic
(N = 241) Cohen's d
WhiteB
(N = 177)
Black
(N = 177) Cohen's d
M SD M SD W/H M SD M SD W/B
Total Hardiness 99.7 14.2 101.4 15.5 -0.12 99.5 14.9 104.9 13.3 -0.38
Challenge 98.9 14.6 102.6 15.1 -0.25 98.8 15.2 106.1 13.5 -0.51
Control 101.8 15.3 101.3 17.0 0.03 101.6 16.0 104.8 15.5 -0.20
Commitment 99.7 14.0 101.3 15.1 -0.11 99.6 14.0 103.8 13.9 -0.30
Notes: WhiteH: Matched sample of White individuals from the norm sample matched to Hispanic
individuals of the norm sample. WhiteB: Matched sample of White individuals from the norm sample
matched to Black individuals of the norm sample. W/H = White vs. Hispanic ratio; W/B = White vs. Black
ratio. Guidelines for evaluating d are 0.20 = small, 0.50 = medium, 0.80 = large.