Download Personality and Insurance Claims PDF

TitlePersonality and Insurance Claims
File Size2.3 MB
Total Pages357
Document Text Contents
Page 2



Content Page

List of Tables 7

List of Figures 8

Abstract 9

Declaration 10

Copyright Statement 11

Dedication 12

Acknowledgements 13

Chapter 1: Introduction 14

1.1 Thesis Overview 14

1.2 Defining the Outcome Variables: Insurance and Credit Usage 15

1.2.1 Insurance 15

1.2.2 Credit 16

1.3 Significance of the Topic 18

1.3.1 Credit and Debt 18

1.3.2 Insurance 19

1.3.3 The Business Case: Should Financial Institutions Care? 21

1.4 A Psychological Approach to Economic Variables 22

1.5 Summary 25

Chapter 2: Personality 27

2.1 Personality 27

2.1.1 What is personality? 28

2.1.2 State, trait and type 29

2.2 Identifying and organising personality traits 32

2.2.1 Five broad factors 33

2.2.2 Five broad factors? 35

2.3 Measurement 37

2.3.1 Personality and CFA 38

2.3.2 Personality measurement is a work in progress 39

2.3.3 If the model doesn’t fit, change the method 40

2.3.4 Summary 41

2.4 Coverage 42

2.4.1 Additional broad measures 42

2.4.2 Specific traits 44

2.5 Prediction 46

2.5.1 Traits beyond the FFM factor space 48

2.5.2 Broad vs Narrow Traits 49

2.5.3 Conclusion 52

2.6 Summary 53

Page 178


results are in line with previous surveys conducted by the ABI (2011) which asked

whether or not people had previously submitted fraudulent claims. The three most

endorsed scenarios were scenario six (claim at every opportunity), scenario one (false

whiplash claim) and scenario five (claim a different driver was driving following an

accident) which were condoned by 145 (30%), 108 (23%) and, 92 (20%) participants

respectively. Scenarios two, three and four was endorsed by between 10% and 15% of

the sample, whilst scenarios seven (friend steals and burns car) and eight (deliberately

cause damage to own property), which are the most clear examples of illegal behaviour,

were condoned by 3% and 6% of the sample respectively.

5.3.3 EFA of Personality

A Kaiser-Meyer-Olkin (KMO) value of .77 and a significant Bartlett's Sphericity

test (χ
= 16725 (2,850), p= 0.001) revealed that both the sampling adequacy and

strength of variable relations were sufficient to justify a factor analysis of all 74

personality items. The items were drawn from 11 scales entitled: Impetuousness, Self-

Regulation, Deferred-Gratification, Consideration of Future Consequences,

Callousness, Conduct-Problems, Integrity, Machiavellianism, Opportunism, Optimism,

and Risk-Taking.

The EFA revealed that 20 factors had eigenvalues greater than unity. The first

ten factors were shown to yield eigenvalues greater than the corresponding factors

suggested by parallel analysis, whilst the scree test indicated discontinuity in the region

of 10 to 12 factors. Thus, factor solutions ranging from 8 to 13 were examined.

The 11, 12 and 13 factor solutions both had factors that were underidentified

(loaded by less than three factor indicators) and thus were discounted. The 8 and 9

factor solutions appeared acceptable but in each solution, one of the factors was

incoherent and didn’t appear to measure a single personality trait. Further, each of these

solutions resulted in the removal of a large proportion of the items (21 and 23

Page 179


respectively). The 10 factor solution, which is discussed below appeared the best on the

basis of the criteria outlined above.

The initial 10 factor solution offered 10 coherent factors and produced the

fewest number (7) of cross factor loadings larger than 0.20. A number of items,

however, failed to load on any factor above the 0.3 cut off. Each of the low loading

items were removed one at a time, and following the removal of each low loading item

the 10 factor solution was re-estimated until the solution satisfied all criteria. The final

10 factor solution accounted for 52.5% of the variance in the data, and retained 62 of the

original 74 items (83%). Thus, the 10 factor solution was deemed to be the best fit to the

data. The pattern matrix generated by the final 10 factor solution is displayed in Table


5.3.4 Factor interpretation

Factor one was predominantly loaded by items from the Ray and Najman

Delayed-Gratification scale and as such I interpret this factor as Deferred-Gratification.

The three highest loading items of the scale refer to the deferment of gratification in a

financial sense. For instance the highest loading item is “I am good at saving money

rather than spending it straight away”.

The second is interpreted as Callousness as it comprises all but one of the items

originally taken from the DAPP-BQ Callousness scale and is supplemented by several

items from other scales that also tap into callous behaviours. A typical item is “I do not

feel guilty when I hurt someone’s feelings”.

Factor three is predominantly loaded by items taken from the IPIP

Machiavellian scale (Goldberg, 1999a). An example item is “I find it easy to

manipulate others”. For these reasons the scale was interpreted as a measure of


Similer Documents