Measuring the impact of the personality trait self-esteem on job attributes and job choice

(1)

Measuring the impact of the personality

trait self-esteem on job attributes and job

choice

By

J.G.J.D. Hoogeboom

(2)

Measuring the impact of the personality

trait self-esteem on job attributes and job

choice

By

J.G.J.D. Hoogeboom

Universityof Groningen Faculty of Economics and Business

MSc Marketing Master Thesis

June 2015 Celebesstraat 23 9715 JB Groningen (06)30374092 S2395924@student.rug.nl Student number 2395924 Supervisor: Dr. F. Eggers

(3)

Management Summary

Whilst most research on job choice focusses on either the post-choice satisfaction phase, or the pre-choice search phase, little research has been done about the actual pre-choice itself. Based on popular studies in the field of job preferences a research has been set up to answer the following question:

How can companies in the Netherlands use personality traits to determine the job preferences of individuals and predict an accurate level of job fit?

Four attributes have been selected to represent the choice set in this research. Salary, opportunities for training, contractual agreement and brand-equity are four attributes which, as proven by academic research, significantly influence the choice process of a consumer. This research aims to measure the relative importance of each attribute and how preferences shift when different levels of each attribute are presented. Next to job attributes, personality traits have shown to be a great influence on the weight of these attributes. Out of several personality traits researched four new hypothesis have been set up stating the influence of the self-esteem trait on the relationship between the attributes and choice.

A design has been set up where salary has four levels, opportunities for training three, contractual agreement four and brand equity three. This gives 3x4x3x4 = 144 different combinations of choice. Ultimately respondents had to make their choice within 10 choice sets where two alternatives were shown and a none-option was added. The survey itself consisted of three parts namely; general questions about the respondent, a validated measure of self-esteem and a choice-based conjoint measure.

(4)

Abstract

Many researchers have attempted to research personal characteristics and their effect on the job search- and choice process. This study measured several job attributes; salary, job security, opportunities for training and brand equity with a moderating role of self-esteem. A model is proposed where all four factors are measured to influence job choice with changing preferences and weights based on self-esteem. Results show that, amongst 63 participants, all four attributes significantly influence the choice outcome of respondents. Furthermore self-esteem positively

influences both the importance of salary and opportunities for training. Indicating different preference levels for respondents with either high- or low self-esteem.

Keywords: Job choice, self-esteem, salary, brand-equity, personality traits, opportunities for training, job tenure.

Research theme: Job Choice Thesis supervisor: Felix Eggers

1. Introduction

Personality traits and job choice are key research concepts within marketing research and practice (Chapman, Uggerslev, Carroll, Piasentin and Jones 2005; Kanfer, Wanberg and Kantrowitz 2001). However most research focuses on the post-choice stadium where different aspects are related to job satisfaction and performance (Harzer and Ruch 2012; Judge and Bono 2001; Jurgensen, 1978). This research intends to bridge the gap and closely examine the choice process and the influence that personality traits have. Specifically, self-esteem is selected as a personality trait to influence the importance of salary, opportunities for training, contractual agreement and brand equity.

(6)

6

information and learning. (Lord, De Vader, and Alliger, 1986). Furthermore these cognitive personality traits affect the extent to which an individual achieves success in life and socioeconomic outcomes (Borghans, Duckworth, Heckman, and Ter Weel, 2008). More research is needed on this concept to link personality traits to job choice. Firstly the job attributes that influence a participant’s choice will be discussed and linked to personality. Next a personality chapter will delve more in the general concept and will specify this construct. Lastly a research question will be proposed with a discussion of managerial and marketing relevance.

1.1 Job attributes

Several changes in job attributes have influenced the job market and its environment over time. Firstly current beliefs exist about the length of job tenure as its slowly decreasing showing signs of so-called “job hopping”. A study done by Hollister and Smith (2014) reveals that the overall average job tenure was at a steady 7.5 years in both 1983 and 2012. When accounting for gender the decrease in tenure for men and increase for women seem to cancel each other out. Further support is gained by revealing that both employers and employees no longer feel obligated to pursue long-term relationships indicating the interest in this attribute.

Secondly, next to decreasing job tenure, a shift in employment conditions is also visible. A slow but steady change from standard working conditions to non-standard working conditions is evident (Hosking and Western, 2008). Nonstandard employment refers to “employment relations that

depart from standard work arrangements in which it was generally expected that work was done full-time, would continue indefinitely, and was performed at the employer’s place of business under the employer’s direction” (Kalleberg, 2000). The result is slightly decreasing job tenure in an

increasingly non-standard work environment.

(7)

7

1.2 Personality

A study done by Harzer and Ruch (2012) indicated that employee’s showed increased signs of happiness when they were able to apply more personal strengths in the workplace environment. These strengths are called signature strengths and are distinctive for the person who uses them (Park, Peterson, and Seligman 2004). Additionally, assumptions are made by several researchers in their studies that job seekers are not always rational in their process of job choice but rather rely on emotions and personality traits (Kidd, 1998; Emmerling and Cherniss, 2003). This calls for more investigation on the personal aspect of the job choice process.

Personality characteristics that focus on the individual can be defined as self-concept. Baumeister (1999) provides the following definition: "the individual's belief about himself or herself, including

the person's attributes and who and what the self is". Judge and Bono (2001) conducted a

meta-analysis on four different personality traits which are self-esteem, self-efficacy, locus of control and emotional stability. These traits correlated with workplace satisfaction and job performance. Results from this study indicated that, based on 274 correlations, these four indicators predict job performance and satisfaction. Amongst these characteristics, self-esteem was found to influence the weight of rational attributes to a great extent (Ellis and Taylor, 1983; Solomom, 1983). Yet their research clearly focuses on job performance and satisfaction after the search process. This research will go one step back and focus on the job choice process. A core personality trait, self-esteem, was found to predict the salary of college students spanning an eight year gap and was three times higher than originally estimated (Drago, 2008). The influence of personality seems to be greater than estimated.

(8)

8

introduction and the theory section, self-esteem is the most interesting personality trait to investigate further. Based on this trait, differences in the job choice process can be determined. A research question can be stated for this paper which is as follows;

How can companies in the Netherlands use personality traits to determine the job preferences of individuals and predict an accurate level of job fit?

This question aims to closely inspect whether people who differ in their personality also make different choices during the choice process, and to what extent. Especially the influence on the changing attributes discussed in the previous chapter is interesting. Based on further theoretical investigation relevant emotional factors or models will be chosen to investigate.

1.3 Marketing / managerial relevance

The preceding paragraphs clearly distinguishes the importance of the individual in the job context. With the results of this study, companies can predict an accurate level of job fit based on the personality traits an employee possesses and whether or not their job choice is a good one. Consumers can use this to determine the best possible career path or direction. Both parties can benefit from reduced post-job dissonance if applied correctly.

This research allows for extension on the job choice process by further investigating the personal reasoning behind job choice instead of rational arguments. Furthermore it can be used to bridge the research gap between current post-choice researches (Harzer and Ruch 2012; Judge and Bono 2001; Jurgensen, 1978) and pre-choice research (Chapman, Uggerslev, Carroll, Piasentin,and Jones 2005; Kanfer, Wanberg and Kantrowitz 2001).

(9)

9

2. Theoretical background

Continuing from the introduction, the job choice and personality traits are the central topics for this thesis. The first paragraphs will conceptualize the job search process and discuss empirical evidence on the topic of job choice. After this the selected personality trait that influences the job choice process will be hypothesized and lastly possible influences on this relationship will be discussed.

2.1 Job search and choice

Various authors have attempted to create definitions of job search behavior and the nature of this phenomenon. Firstly Schwab, Rynes and Aldag (1987) proposed a model where job search behavior was split into the sources used to acquire information as well as the intensity and extent to which these were pursued. Blau (1994) stated that the job search process consisted of a preparatory phase and an active phase. The same year Barber, Daly, Diannantonio and Philips (1994) suggested job search behavior points to acknowledging the existence of job opportunities and acquiring more details on job alternatives. All these models and definitions have a similar nature and that is that job choice is multidimensional. No single aspect is responsible for the outcome during the job choice process. Different attributes, already discussed during the introduction, are influencing the decision making process.

Regarding literature about job choice there is still much to learn. Current literature mostly focuses on a few if not a single aspect to determine job choice. This paragraph aims to break down different literature studies and find important determinants to measure job choice.

It is vital for an organization to attract qualified employees in order to gain a competitive advantage. Yet, little is known about the specific characteristics that drive individuals in selecting a position. More detail is needed on these characteristics and the interdependency that comes with them. Four relevant attributes have been determined for job choice which will be discussed below.

Salary; Firstly Cable and Graham (2000) have shown that pay level influences perceived employer

(10)

10

not only on personal work values but also significantly on pay and advancement potential which signifies the next attribute. Based on these studies salary needs to be added as a control variable

Opportunities for training; Secondly Kaliprasad (2006) researched that training opportunities at an

organization significantly influences the capability of an organization to attract better qualified staff. Organizations should apply effective strategies to allow employees to participate in key organizational dimensions and providing them with resources for growth and learning. Chapman et al. (2005) found that the job-pursuit intentions amongst other predictors are capable of influencing the decision process of an applicant.

Job tenure; Third Jurgensen (1978) did a dated, but effective research on job tenure, job

advancement and pay which yielded profound results. A scale could be made for different genders what participants found important about their job. For men the order was security, advancement, type of work, company, pay, co-workers, supervisor, benefits, hours, and working conditions. The previous two attributes, salary and opportunities for training, are also included in this research. More recent research by Kalleberg (2000) and Hollister and Smith (2014) showed both job tenure and employment conditions change requiring different strategies for attracting and retaining capable staff.

Brand-equity; Lastly DelVecchio, Jarvis, Klink and Dineen (2007) linked organizations with

high-brand equity to be more favorable amongst applicants. Keller’s definition of brand equity is “the

differential impact of brand knowledge on consumer response to the marketing of the brand”. One

of the most important returns that an organization can gain from having a strong brand is the ability to charge higher prices. Consumers are willing to pay premium prices for products even though the less branded products are of the same quality (Park and Srinivasan, 1994). The same concept can be applied in the job choice context where individuals are willing to forfeit certain benefits to be affiliated with a strong brand and thus, increasing the relative importance of this attribute.

(11)

11

Table 1: Overview of core studies

Author Key dimensions, sample and findings Basis

Cable and Graham, 2000

A study in which the pay level is one of the used variables to measure perceived employer reputation and interest.

H1

Kaliprasad, 2006 Some highly effective strategies to attract capable staff consists of in-house career development and opportunities for promotion

H1

Jurgensen, 1978 57,000 job applicants ranked the importance of factors that make a job good or bad. Security, advancement, type of work, company and pay are a few of the significant predictors

H1

DelVecchio et al, 2007

A study amongst 385 participants showed that job opportunities with strong brands are more favorable.

H1

Judge and Bono, 2001.

Core self-evaluation traits influence job satisfaction and performance.

H2, H3

Judge and Cable, 1997

A study of 182 participants found that the subjective fit was related to perceived organizational attractiveness and that personality traits influenced workplace preferences.

H2, H4

Ellis and Taylor, 1983

A study amongst 86 participants found that self-esteem predicted the length of intended tenure along with other factors.

H3

Solomon, 1983 Research showed that consumers used brand-equity to enhance their personal identity (self-esteem) and status.

H4

Waters., McCabe, Kiellerup and Kiellerup, 2002

A study amongst 77 respondents about mentoring, career-related support and esteem found significant evidence relating self-esteem to support and frequency of contact.

H5

Lent, Brown and Larkin, 1986.

A study of 105 participants shows self-efficacy directly affects the perceived range of career opportunities

H5

Note. Additional studies were used to solidify the propositions for H2-H5. See the appropriate chapter for clarification.

(12)

12

H1: Higher salary, better job security, more opportunities for training and better employer-based brand equity make a job offer more desirable.

Salary and job security are measurable by hard numbers about paid salary and contractual period. Opportunities for training and employer-based brand equity have to be added story wise. The next step in this research is investigating how personality affects the decision making process within the job choice context.

2.2 Personality; Self-esteem

The influence of personality during a job decision making process remains mostly unknown. Most research concentrates on the post-choice effects of personality differences and emotional stability (Harzer and Ruch 2012; Judge and Bono 2001; Jurgensen, 1978). Neglecting personality or emotional stability during the choice process might lead to an incomplete understanding of why certain choices are made. The following paragraphs will summarize the studies who did focus on the decision making process.

Most job searches and choices contain an emotional aspect. This refers to the study done by Bonaccio, Gauvin and Reeve (2014) who research choice strategies. Their result showed that emotional language during the job search and choice process resulted in a less criterion driven strategy. One of the emotional constructs that consolidates different personality aspects is called self-concept (Rosenberg, 1965; Judge and Bono, 2001).

From the introduction it is clear that the rational aspects of job choice are being turned over by the irrational or emotional factors. The emotional factors examined by Judge and Bono (2001) will be further discussed.

Their research focused on four emotional traits which are part of the general construct called self-concept as defined in the introduction (Baumeister, 1999). The four personality traits measured by Judge and Bono (2001) are self-esteem, self-efficacy, locus of control and emotional stability which will be defined further. Self-esteem can be defined as “A reflection of a person’s overall

evaluation or appraisal of his or her own worth”. Darley (1999) found that individuals with high

(13)

13

spending. The next factor is generalized self-efficacy which can be defined as “one's estimate of

one's fundamental ability to cope, perform, and be successful”. The third factor is locus of control

which is “a manifestation of core evaluations because internals believe they can control a broad

array of factors in their lives”. Lastly there is emotional stability which is “reflecting the tendency to be confident, secure, and steady” (Judge and Bono, 2001). Results from this study found that all

four factors positively correlated with job satisfaction and performance. The previously mentioned research by Bonaccio, Gauvin and Reeve (2014) showed emotional language was negatively related to salary. A strong emotional stability might overturn this effect.

Alavi and Askaripur (2003) examined multiple factors and their relationship with self-esteem in a governmental setting. Positive effects were found between self-esteem and job satisfaction from wages, co-workers and management. A more recent study by Drago (2008) implied that the effects of self-esteem on salary were two to three times higher than the research done by the National Longitudinal Survey of Youth originally estimated. Using the Rosenberg scale for self-esteem an accurate prediction could be made what college students would earn eight years later.

These studies found self-esteem to be an accurate predictor on the choice and decision making process during job search. Examining self-esteem as a specific approach to personality is chosen based on the findings of the authors discussed in the previous paragraph. The resulting hypothesis for the first moderation effect can be defined:

H2: As self-esteem increases, a consumer places more value on salary, and hence is more responsive to salary, thus increasing the effect of salary on job choice.

Ellis and Taylor (1983) have also done notable research on specific attributes of personality. During a four-month study they found that self-esteem predicted the following: The sources used, the job offers received, the evaluations from organizational recruiters, satisfaction during the search process, the amount of offers received, length of job tenure and lastly the acceptance of jobs before graduation. These findings can be summarized in three ways.

(14)

self-14

esteem affected the applicant’s social skills which resulted in fewer job offers on the long term (Ellis and Taylor, 1983). Participants with high self-esteem receive more job offers and it can be expected that job security matters less since high self-esteem respondents are having less trouble finding good jobs. A hypothesis can be created based on previous research:

H3: As self-esteem increases, a consumer places less value on job security, and hence is less responsive to job security, thus decreasing the effect of job security on job choice.

Not only salary and job security are influenced by the perceived effect of self-esteem. Self-esteem can directly affect the extent of the influence that brand-equity has on an applicant. Tavassoli, Sorescu and Chandy (2014) researched that brand-equity influences employee’s choices. Employees would forfeit on other attributes (e.g. salary) to attain a position at a high brand-equity organization. Judge and Cable (1997) examined the influence of personality on workplace preferences relating to culture, objective and subjective fit. The applicant’s personality traits and direct perception of fit within an organization was found to be a significant predictor to organizational attractiveness. Lastly Solomon (1983) showed that individuals used brand-equity to enhance and express their identity both publicly and privately (e.g. self-esteem and status). Being able to work at Apple can accord to an employee’s perception of self-esteem and status or enhance the already existing perception of it. Thus self-esteem is estimated to positively affect the relationship between brand-equity and job choice. The resulting hypothesis is as follows:

H4: As self-esteem increases, a consumer places more value on brand-equity, and hence is more responsive to the brand-equity of an organization, thus increasing the effect of brand-equity on

job choice.

(15)

15

support and frequency of contact. This would imply for this study that high self-esteem respondents value mentor support more. Lastly from the works of Oakes, Ferris, Martocchio, Buckley and Broach (2001) it is evident that personality traits positively correlate with skill acquisition which in turn predicts the level of job performance. One of the personality traits used here was self-esteem and individuals with high self-esteem were more open to skill acquisition. It can therefore be expected that individuals with higher self-esteem value training opportunities more than individuals with low self-esteem. The last hypothesis can be derived from this:

H5: As self-esteem increases, a consumer places more value on training opportunities, and hence is more responsive to opportunities for training, thus increasing the effect of training

opportunities on job choice.

Concluding the previous chapters it is apparent that self-esteem can affect the strength of the proposed attributes and that it should be included in the conceptual model. Based on the hypotheses a conceptual model can be drawn.

Figure 1: Conceptual model

(16)

16

3. Methodology

In this chapter the methodology part of the research will be clarified. Firstly, sampling and procedure and analysis will provide a justification for the method of analysis. The attributes and levels will further specify the survey set-up. Lastly the experimental design and model propose a measurable set-up for analyzing the results.

3.1 Sampling

Data will be gathered by means of a survey conducted amongst Dutch respondents. Dutch respondents have been chosen since these depict the Dutch market and can be gathered from the networks of friends and personal contacts. A minimum sample size of 50 respondents is required to provide accurate measurements. Through usage of a personal network it is expected to receive the most responses from either students or working respondents.

The survey will be conducted online and will be open to any respondent within the Netherlands. There will be no focus on any specific group in this study (e.g. students, working people).

3.2 Procedure and measures

The survey itself will consist of three parts. The first part will have some general questions about the respondent (e.g. age and education) to get insight in the demographic data of the respondents. This makes the research more accessible and reproducible for other researchers. The second part of the research measures the respondent’s level of self-esteem using the Rosenberg’s self-esteem scale (1965). This questionnaire is a validated scale to measure the construct self-esteem and was developed by Dr. Morris Rosenberg. The measurement consists of ten different questions that allow for accurate measurement of a respondents level of self-esteem. Participants can choose from: Strongly disagree, disagree, neutral, agree or strongly agree. To catch participants trying to rush through the survey different questions are reversed in valence. Participants scoring high on questions 1,2,4,6 and 7 should score low on items 3,5,8,9 and 10 since the questions are asked in a reverse way. A copy of the questionnaire can be found in Appendix A and B.

(17)

17

the analysis and a solid base can be generated for predicting consumer behavior (Eggers and Sattler 2011). This part of the survey will be a choice conjoint analysis showing job choices with different levels of salary, opportunities for training, job security and brand-equity. The no-choice option exists to not force respondents into choices they would otherwise not agree with and keep the representativeness of the survey adequate. No-choice also provides the option to test whether the relationship between attributes are positive or negative.

Ethics and assurance; respecting the privacy of the individual taking the test should be prioritized.

An agreement will be set up for participants who start the test introducing the questionnaire and explaining what will happen. Respondents will be reassured that their data will be carefully documented and not handed out to any third party.

3.3 Attributes and levels

Four different attributes have been selected in this research to form the basis of a choice-based conjoint analysis; salary, opportunities for training, contractual agreement and brand equity. Salary is represented by four levels of monthly income based on the average Dutch income of 2.734 euro (based on CBP forecast of March 2015) in 2015 if 2,734 euro. Opportunities for training is represented by a waiting period before opportunities for training and advancement are available. Contractual agreement uses the main forms of contract in the Dutch market and finally brand equity is represented by company size and brand. No specific companies have been chosen to represent the brand equity attribute since respondents each have a different subjective opinion about brand equity organizations. The resulting attributes and levels have been provided in an overview in table 3.1.

Table 3.1: Attributes and levels regarding the research design

Attribute Levels

Salary 2500

2650 2800 2950

(18)

18

There are possibilities for growth and training after 3 years Contractual It is a temporary contract

It is a contract for 6 months It is a yearly contract It is a two-year contract

Brand equity You will work at a small unknown company with an unknown brand and only a few employees present

You will work in a medium-sized organization with a locally known brand and multiple establishments in the region You will work at a large multinational with a well-known brand and many establishments and employees around the world.

Note: the questionnaire showed the brand-equity level as “company size and branding” since brand-equity is not a common used English word and could be misinterpreted.

3.4 Experimental design

The design of this study consists of all the levels used in table 3.1. Salary and contractual agreement have four levels and opportunities for training and brand-equity have three levels. Therefore, 4x4x3x3=144 different profiles could be created. However respondents are not asked to choose between all 144 profiles. In total 10 different choice sets have been selected for participants to complete. More than 12-15 choice sets would result in a negative impact due to increased boredom and decreasing attention (Eggers and Sattler, 2009).

(19)

19 Figure 3.1. Example choice set my.preferencelab.com

Note: all actual choice sets can be found in appendix B:

In this design the type of job is referred to as general. Specific jobs can influence the relative importance each attribute has. In order to keep this research neutral this text has been added above every choice set.

3.5 Model

The resulting analysis will reveal utility values for each of the different attributes. The measurement of utility is done through the following formula:

𝑉𝑛𝑖= ∑ 𝐵𝑛𝑘∗ 𝑋𝑖𝑘 𝐾

𝑘=1

(20)

20

4. Results

This result section will discuss the findings of the online survey. The first subsection gives an overview of the steps taken before the results were ready for analyzing. The second subsection will discuss the descriptive statistics. The third and fourth subsection will present the results of the main relationship and predictive validity statistics. Lastly subsection five will assess the moderator hypotheses.

4.1 Sample

In total 63 respondents completed the questionnaire. After removing the two respondents mentioned earlier a total of 61 were used for the final analysis. The average age of these respondents was 26. Over two-thirds of the respondents were male (42 male, 20 female). Lastly the educational status of the respondent was as follows; 20 respondents were still students, 41 respondents were working, 0 were job-seekers and 1 person responded with none of the above.

4.2 Reliability

Before the initial result scan several steps need to be taken to ensure a reliable dataset for choice-based conjoint analysis. Firstly respondentID 113 and 239 were removed from the dataset for answering “none” at all the options and thus rushing through the survey. No other respondents had to be removed after scanning for outliers and time used to complete the survey.

Secondly recoding within the dataset had to be done for the moderator variable self-esteem. Using Rosenberg’s self-esteem scale five out of ten questions were reversed in valence. These were recoded and a Cronbach’s alpha analysis was done to ensure combining the variables was possible. The results of this analysis are shown in table 4.1

Table 4.1. Cronbach’s alpha statistics for self-esteem

Question Cronbach’s alpha if item deleted

Question Cronbach’s alpha if item deleted

Esteem1 .855 Esteem6 .853

(21)

21

Note: The total Cronbach’s alpha value for all 10 items was .876

The baseline value for Cronbach’s alpha is higher than .65 (Malhotra, 2007). The total resulting value for self-esteem was .876. Removing any items would not increase the Cronbach’s alpha any further. All ten questions were used to measure the respondent’s average self-esteem value.

Lastly some final recoding in the data had to be done. Since a no-choice option has been added an extra zero is present in all attributes. This zero needed to be recoded to system-missing values. Salary was also recoded to show the actual salary values instead of option 1, 2, 3 and 4. Using the actual values would allow salary to be a linear attribute that can be extrapolated for better information and prediction purposes.

4.3 Main effects Model

Assessing the main effect of all four attributes on job choice will consist of an analysis of fit, in which model properties will be determined and the overall fit value of the models will be discussed. Secondly the main relationship utility values will be discussed after which the first hypothesis can be answered.

4.3.1 Fit values

Before the main relationship can be assessed the fit values will be discussed for two possible models. The difference between these models is that Salary can be a linear attribute and both models should be tested (one with salary as a linear attribute and one with salary as a part-worth attribute). The utility range for salary is shown in figure 4.1 and linearizing is possible. Table 4.2 depicts the Fit values needed to conduct a chi-square test for model difference. The best model will be used for further analysis. The other model will be shown in appendix C for reference.

Predictive validity; the predictive validity can be calculated manually by usage of the hit-rate box.

(22)

22

validity of this model is much better than the native model which has a predictive validity of 33.34. It can be concluded that the predictive validity of both models is adequate.

The possibility of linearizing salary can also be viewed in figure 4.1. This figure demonstrates that salary indeed has the option of linearizing. Even though the last attribute level (2950 euro i.e. case number 4) has a slightly receding line the model as a whole can be linearized.

Figure 4.1: Utility values for part-worth salary (Case no. 1-4 = Salary levels 2500-2950)

The fit values can be concluded in table 4.2 where the log likelihoods, parameters, R-squared, adjusted R-squared and hit rate have been added for both models.

Table 4.2. Linear versus part-worth model

LL(B*) LL(0) Parameters R-squared R-adj Hit rate

Linear -425.5553 -659.17 9 0.4884 0.4776 66.72%

Part-worth -424.2320 -659.17 11 0.4900 0.4767 66.89%

Note: The final choice for model was based on R-adj.

(23)

23

part-worth having comparable better outcomes. When adjusting for the amount of parameters however, the adjusted R-squared of the linear model is higher. Usage of this model results in 47.76% explained variance. To calculate whether these models differ a chi-square test is used.

Chi-square; the chi-square statistic for measuring difference is determined by the following

calculation using the LL(B*) from both models: -2((-425.5553)-(-424.2320)) = 2.6466 wit 2 degrees of freedom returns a value between (p = 0.9) and (p = 0.1) in the chi-square table. From this statistic it can be concluded that the models do not significantly differ from each other. Based on this result a linear model will be used for further analysis since it is more parsimonious. Advantages of this approach is that extrapolation of salary is now possible and will be done in later stages.

4.3.2 Estimates

Based on the hypothesis and a linear model the main relationship will be assessed first. This analysis consists of a partly linear model as determined in paragraph 4.3. The model will be used to determine the attribute importance, overall significance and utility values. Based on these utility values a model will be proposed to answer the hypothesis.

Starting with the first step, all estimations of the model are done on a part-worth level apart from Salary. The part-worth are being estimated in order to determine the relative importance of the attributes for job choice. An overview of the estimated utility values, p-values, ranges and relative importance of the attributes is shown in table 4.3.

Table 4.3 Utility levels for a salary-linear model

Attribute Level Estimated utility value

(24)

24 2 years 0.4490 Equity Small -0.2362 19.0471 <0.0001 0.5689 17.85% Medium 0.3327 High -0.0865 None opt. 2.4972 6.7055 0.0096 2.7104 /*

Note: The R-squared for this model is 0.3512

This model reveals different significance statistics than the previous estimation shown in Appendix C. Both Salary and none-option became less significant (p = 0.0019 and p = 0.0096) but still well within the marge (p > 0.05). A new possibility with a linear effect is extrapolation. Salary levels outside the four levels can be calculated. A salary of 2000 reveals a total utility value of 0.0019 * 2000 = 3.80. Now that the utility values have been estimated, choices can be predicted based on these values versus the added none-option.

Where the probability of a certain choice is the sum of the exponential utility values divided by the sum of exponential utility values plus the exponential value of the none-option. For example an offer has been placed with a salary of 2650, with opportunities for training after 2 years, a 1 year contractual period with a low brand-equity company. The resulting formula would be:

𝑃𝑟𝑜𝑏𝐶_𝑖 = ∑ 𝐸𝑥𝑝((2650 ∗ 0.0019) − 0.0515 + 0.2294 − 0.2362)

∑ 𝐸𝑥𝑝(((2650 ∗ 0.0019) − 0.0515 + 0.2294 − 0.2362) + 𝐸𝑥𝑝(2.4972) The above formula results in a probability value of 0.91 indicating that respondents would be 91% likely to choose this job over the no-choice option. Reducing the salary to 1500 euro would make this offer 50% likely to be accepted surpassing the so called “willingness to pay” border. The best utility values, as seen in table 4.3 are training opportunities after 1 year, a two year contract and a medium brand equity company. The most attractive salary in this set was 2950 euro. Combining all these utility values results in 98% acceptance chance.

(25)

25

4.4 Moderation effect

There are four hypotheses that state a moderating effect exists between self-esteem and the four attributes. Before the analysis of utility values a fit test will be done to compare the models. Then, based on a complete model, each moderating effect will be discussed separately.

4.4.1 Fit Values

Firstly the fit values of the moderated model versus the linear model will be presented. The fit values can be concluded in table 4.4 where the log likelihoods, parameters, squared, adjusted R-squared and hit rate have been added for both models.

Table 4.4. Linear versus part-worth model

LL(B*) LL(0) Parameters R-squared R-adj Hit rate Linear vs native -425.5553 -659.1700 9 0.4884 0.4776 66.72% Moderated vs native -422.5078 -659.1700 17 0.4897 0.4693 66.89% Moderated vs linear -422.5078 -425.5553 17 0.0072 -0.0328 66.89% Note: The final choice for model was based on R-adj.

The moderated model will first be compared to the actual native model. With the increase of the amount of parameters the R-squared increases slightly from 0.4884 to 0.4897. However when adjusting for the amount of parameters the adjusted R-squared is lower.

When comparing the moderated model to the linear model the R-squared and adjusted R-squared column provide the increase or decrease in value. The R-squared increases with 0.0072. However, the adjusted R-squared decreases by 0.0328. To test whether the models differ significantly a chi-square test is used.

Chi-square; the chi-square statistic for measuring difference is determined by the following

(26)

26

4.4.2 Estimates

As expected in hypothesis two through five, self-esteem has a significant influence on the effect of the four attributes. For each of the attributes the relevant moderation model will be presented. Table 4.4 presents the estimated utility values and p-values including moderation.

Table 4.4. Main attributes and levels accounting for moderation

Attribute Level Utility value Wald P-value

Salary Linear 0.0018 24.7378 <0.0001 Training 1 year -0.9214 6.8331 0.033 2 years 0.8455 3 years 0.0759 Contractual Part-time -0.8677 5.9642 0.11 6 months -0.5555 1 year 0.627 2 years 0.7962 Equity Small 0.5017 5.3515 0.069 Medium 0.3857 High -0.8874 None-option Effect-coded -2.8814 8.2985 0.0040

Salary x Esteem Linear 0.0005 5.1449 0.0212

Training x Esteem After 1 year 0.5009 12.8142 0.00034

After 2 years -0.3144 5.9505 0.015

After 3 years -0.1864 Reference level

Contractual x Esteem Temporary 0.1325 0.6314 0.43

Six-months 0.1087 0.4022 0.53

One-year -0.1294 0.5223 0.47

Two-years -0.1118 Reference level

Equity x Esteem Low-equity -0.2676 4.3942 0.036

Med-equity -0.0122 0.0078 0.93

High-equity 0.2798 Reference level

(27)

27

Both salary and opportunities for training have a significant effect in this model (p < 0.05). Brand equity is very close to significant (p = 0.069) and contractual agreement is not significant (p = 0.11). The moderation effects will be discussed next.

Table 4.4 depicts two levels of training to be significant. LatentGold presents the moderation effect based on the reference levels of the attributes. However the amount of reference levels is the amount of attribute levels minus one. The last level needs to be recovered and calculated. These calculated utility values are multiplied with the level of self-esteem to reveal the actual influence. Respondents varied in esteem on a continued scale from low esteem (0) to high self-esteem (4).

Salary; the linear attribute Salary has only one utility value and moderation effect. The baseline

utility value is 0.0018 which is only slightly different from the 0.0019 from the model without self-esteem in table 4.3. However the moderation effect needs to be taken into account which is 0.0005 times the level of self-esteem. The resulting formula becomes (0.0018 + 0.0005( * Self-esteem)) * Salary. From the moderation effect it is clear that a higher self-esteem adds value to the original estimate increasing the utility value. These results are in line with the second hypothesis that states; as self-esteem increases, a consumer places more value on salary, and hence is more responsive to salary, thus increasing the effect of salary on job choice.

For the second hypothesis H0 can be rejected in favor of the alternative. Self-esteem significantly moderates the relationship between salary and job choice. Consumers place more value on salary as self-esteem increases and hence, are more responsive to this attribute.

Job security; from the utility values in table 4.4 it is clear consumers place less value on job security

as esteem increases. The formula to calculate the total utility values based on the level of self-esteem is as follows.

- Part-time contract = -0.8677 + 0.1325( * self-esteem level) - Six month contract = -0.5555+ 0.1087( * self-esteem level) - One year contract = 0.627 - 0.1294( * self-esteem level) - Two year contract = 0.7962 - 0.1118( * self-esteem level)

(28)

28

that, as self-esteem increases, a consumer places less value on job security, and hence is less responsive to job security, thus decreasing the effect of job security on job choice. However the p-values for this hypothesis are not significant (p = 0.43, p = 0.53 and p = 0.47). Hence we cannot reject H0 in favor of the alternative.

Brand equity; Brand equity is very close to significant in table 4.4. The fourth hypothesis states

that as self-esteem increases, a consumer places more value on brand-equity, and hence is more responsive to the brand-equity of an organization, thus increasing the effect of brand-equity on job choice.

The utility values in the moderation table match the expected relationship. As the level of self-esteem increases the respondent places more value on brand-equity.

- Small brand equity = 0.5017 - 0.2676( * self-esteem level) - Medium brand equity = 0.3857 - 0.0122( * self-esteem level) - Large brand equity = - 0.8874 + 0.2798( * self-esteem level)

However only the first level of brand equity can be deemed significant (p = 0.036). The influence of self-esteem on the second level of brand equity (medium) is not significantly different from 0 (p = 0.93). This high p-value indicates that it is actually very close to 0. Based on this information H0 for the fourth hypothesis cannot be accepted in favor of the alternative.

Training; The fifth and last hypothesis stated that, as self-esteem increases, a consumer places more

value on training opportunities, and hence is more responsive to opportunities for training, thus increasing the effect of training opportunities on job choice. The calculated utility formulas are the following:

- Training after 1 year = -0.9214 + 0.5009( * self-esteem level) - Training after 2 years = 0.8455 – 0.3144( * self-esteem level) - Training after 3 years = 0.0759 - 0.1864(* self-esteem level)

(29)

29

Based on the preceding analysis section, hypothesis one can be partly accepted. Hypothesis two and five can be accepted in favor of the alternative. Hypothesis three and four showed the correct signs and expectations but were statistically insignificant. An overview of the hypotheses and support can be found in table 4.5

Table 4.5: Hypotheses overview

Hypothesis Support

H1: Higher salary, better job security, more opportunities for training and

better employer-based brand equity make a job offer more desirable.

Partly supported

H2: As self-esteem increases, a consumer places more value on salary, and

hence is more responsive to salary, thus increasing the effect of salary on job choice.

Supported

H3: As self-esteem increases, a consumer places less value on job security,

and hence is less responsive to job security, thus decreasing the effect of job security on job choice.

Not supported

H4: As self-esteem increases, a consumer places more value on brand-equity,

and hence is more responsive to the brand-equity of an organization, thus increasing the effect of brand-equity on job choice.

Not supported

H5: As self-esteem increases, a consumer places more value on training

opportunities, and hence is more responsive to opportunities for training, thus increasing the effect of training opportunities on job choice.

Supported

Note: the data from H3 and H4 followed the predicted direction, but were not significant

(30)

30

5. Conclusion and Discussion

This research initially stated a research question that aims to predict job preferences of individuals based on their personality traits. This question has been further specified to measure the level of self-esteem to influence the four job attributes salary, opportunities for training, contractual agreement and brand equity which is combined in a model after specifying hypothesis concerning each of these relationships.

The results indicate that hypothesis one can be partly accepted and hypothesis two and five can be accepted. Hypothesis one states that higher salary, better job security, more opportunities for training and better employer-based brand equity make a job offer more desirable. The most important attributes were training, contractual agreement and salary on an almost equal level. Brand equity was not as important and did not exhibit a positive influence. Hypothesis two and five specified the moderation influence of salary and opportunities for training which both showed strong significance. These findings are strongly in line with the research of Cable and Graham (2000), Kaliprasad (2006) and Jurgenssen (1978). The research question “How can companies in

the Netherlands use personality traits to determine the job preferences of individuals and predict an accurate level of job fit?” can be answered by stating that self-esteem can be correctly used to

measure the positive change in influence of salary and training opportunities.

Hypothesis three and four can be rejected since they did not provide significant results. However, for brand equity interesting results were found. Although most attributes are part-worth, the brand equity attribute seems to exhibit a quadratic relationship because of decreasing returns. Respondents seem to prefer companies with a medium level of brand equity over those with high brand-equity. This creates new questions based on the research of Tavassoli, Sorescu, Chandy (2014), Solomon (1983) and Delvecchio et al. (2007) who predicted that that respondents favor high brand-equity companies. More investigation is needed examining this quadratic relationship possibly extending the research of Delvecchio et al. (2007).

Discussion: There are good possibilities for future research based on this thesis. Firstly this conjoint

(31)

31

this research can be expanded to other personality traits and determine which trait causes changes in attribute preferences.

The survey itself consisted of three choices; two choices with attributes and a none-option. The none-option statistics cannot be blindly trusted however. Most respondents chose between the two choice options and would only pick none if both options seemed equally attractive or unattractive. One method to fix this is to make respondents choose between two choice options and ask the question “would you actually take this job if it was offered” afterwards. This would give a more representative utility value for the none-option.

The moderation effect of the none-option might be interesting to discuss too. The none-option provided a positive utility value (the higher a respondents self-esteem, the more weight the none-option has) but these results are not significant (p = 0.073) and thus no conclusion can be drawn from these statistics.

Managerial and Theoretical implications: This research bridges the gap between current

post-choice researches (Harzer and Ruch 2012; Judge and Bono 2001; Jurgensen, 1978) and pre-post-choice research (Chapman, Uggerslev, Carroll, Piasentin,and Jones 2005; Kanfer, Wanberg and Kantrowitz 2001). This model is only a basic model for job attribute preferences. One of the largest studies about job preferences is done by Jurgenssen (1978) but currently getting dated. Based on these implications I would advise to reinvestigate the Jurgensen study in current economic context. This would provide a solid base to link pre-choice and post-choice studies and can be extended in a multitude of ways. For example by adding different personality traits, preferences or changing environmental attributes.

(32)

32

Limitation: One of the limitations of this research is that brand-equity is not specified to a company

or organization. In order to incorporate brand-equity as an attribute it has been changed to company size and branding with a general text. It is expected that the English level of respondents might not be sufficient to understand the concept “equity” and hence influence results. Brand-equity is the only attribute with a quadratic relationship contradicting the research of Delvecchio et al. (2007). Secondly the sample size can be considered narrow. This research can be done with a minimal respondent set of 50. During this research 61 respondents were actually used for the final analysis. Although the respondent count is adequate for research purposes a higher sample size should be considered when conducting future research. Lastly the respondent base consisted mostly of friends, family and employees from the Netherlands. Different cultures or preferences can exists even within the same country which in turn generates a representative research limited to the Netherlands.

Further research can be done by specifying companies to represent the brand-equity attribute. These companies should be validated to ensure a correct representation of the level of brand-equity. Moreover this research focusses solely on self-esteem using Rosenberg’s scale. The current definition of self-concept embodies four personality traits as researched by Judge and Bono (2001). These traits are self-esteem, self-efficacy, locus of control and emotional stability. I would advise to test these personality traits too and how these influence the weight of different job attributes. But not only Judge and Bono (2001) have done notable research on personality. Many other researches that fall outside the range of this paper also discuss personality as a construct (e.g. The Big Five).

(33)

33

References

Alavi, H. R., and Askaripur, M. R. (2003). The relationship between self-esteem and job satisfaction of personnel in government organizations. Public personnel management,

32(4), 591-600.

Barber, A. E., Daly, C. L., Giannantonio, C. M., and Phillips, J. M. (1994). Job search activities: An examination of changes over time. Personnel Psychology, 47(4), 739-766.

Blau, G. (1994). Testing a two-dimensional measure of job search behavior. Organizational

behavior and human decision processes, 59(2), 288-312.

Borghans, L., Duckworth, A. L., Heckman, J. J., and Ter Weel, B. (2008). The economics and psychology of personality traits. Journal of Human Resources, 43(4), 972-1059.

Bonaccio, S., Gauvin, N., and Reeve, C. L. (2014). The Experience of Emotions during the Job Search and Choice Process among Novice Job Seekers. Journal of Career Development,

41(3), 237-257.

Cable, D. M., and Graham, M. E. (2000). The determinants of job seekers' reputation perceptions.

Journal of Organizational Behavior, 21(8), 929-947.

Chapman, D. S., Uggerslev, K. L., Carroll, S. A., Piasentin, K. A., and Jones, D. A. (2005). Applicant attraction to organizations and job choice: a meta-analytic review of the correlates of recruiting outcomes. Journal of applied psychology, 90(5), 928.

Darley, W. (1999). The relationship of antecedents of search and self-esteem to adolescent search effort and perceived product knowledge. Psychology and Marketing,16 (5), 409-427. DelVecchio, D., Jarvis, C. B., Klink, R. R., and Dineen, B. R. (2007). Leveraging brand equity to

attract human capital. Marketing Letters, 18(3), 149-164.

Drago, F. (2011). Self-esteem and earnings. Journal of Economic Psychology, 32(3), 480-488. Eggers, F., and Sattler, H. (2011). Preference Measurement with Conjoint Analysis. Overview of

State-of-the-Art Approaches and Recent Developments. GfK Marketing Intelligence

Review, 3(1), 36-47.

Ellis, R. A., and Taylor, M. S. (1983). Role of Self-Esteem within the Job Search Process. Journal

of Applied Psychology, 68(4), 632-640.

Emmerling, R. J., and Cherniss, C. (2003). Emotional intelligence and the career choice process.

(34)

34

Harzer, C., and Ruch, W. (2012). When the job is a calling: The role of applying one's signature strengths at work. The Journal of Positive Psychology, 7(5), 362-371.

Hollister, M. N., and Smith, K. E. (2014). Unmasking the conflicting trends in job tenure by gender in the United States, 1983–2008. American Sociological Review, 79(1), 159-181.

Hosking, A., and Western, M. (2008). The effects of non-standard employment on work—family conflict. Journal of Sociology, 44(1), 5-27.

Judge, T. A., and Bono, J. E. (2001). Relationship of core self-evaluations traits—self-esteem, generalized self-efficacy, locus of control, and emotional stability—with job satisfaction and job performance: A meta-analysis. Journal of applied Psychology, 86(1), 80.

Judge, T. A., and Cable, D. M. (1997). Applicant personality, organizational culture, and organization attraction. Personnel Psychology, 50(2), 359-394.

Judge, T. A., and Bretz, R. D. (1992). Effects of work values on job choice decisions. Journal of

applied psychology, 77(3), 261.

Jurgensen, C. E. (1978). Job preferences (What makes a job good or bad?). Journal of Applied

psychology, 63(3), 267.

Kaliprasad, M. (2006). The human factor. I: Attracting, retaining, and motivating capable people.

Cost Engineering, 48(6), 20-26.

Kalleberg, A. L. (2000). Nonstandard employment relations: Part-time, temporary and contract work. Annual review of sociology, 341-365.

Kanfer, R., Wanberg, C. R., and Kantrowitz, T. M. (2001). Job search and employment: A personality–motivational analysis and meta-analytic review. Journal of Applied

Psychology, 86(5), 837.

Kidd, J. M. (1998). Emotion: An absent presence in career theory. Journal of Vocational Behavior,

52(3), 275-288.

Lent, R. W., Brown, S. D., and Larkin, K. C. (1986). Self-efficacy in the prediction of academic performance and perceived career options. Journal of counseling psychology, 33(3), 265. Lord, R. G., De Vader, C. L., and Alliger, G. M. (1986). A meta-analysis of the relation between

personality traits and leadership perceptions: An application of validity generalization procedures. Journal of Applied Psychology, 71(3), 402.

Malhotra, N., Hall, J., Shaw, M., & Oppenheim, P. (2007). Essentials of Marketing Research, An

(35)

35

Oakes, D. W., Ferris, G. R., Martocchio, J. J., Buckley, M. R., and Broach, D. (2001). Cognitive ability and personality predictors of training program skill acquisition and job performance.

Journal of business and psychology, 15(4), 523-548.

Park, C. S., and Srinivasan, V. (1994). A Survey-Based Method for Measuring and Understanding Brand Equity and Its Extendibility. Journal of Marketing Research (JMR), 31(2), 271-288. Park, N., Peterson, C., and Seligman, M. E. (2004). Strengths of character and well-being. Journal

of social and Clinical Psychology, 23(5), 603-619.

Schwab, D. P., Rynes, S. L., and Aldag, R. J. (1987). Theories and research on job search and choice. Research in personnel and human resources management, 5(1), 129-166.

Solomon, Michael R. (1983), “The Role of Products as Social Stimuli: A Symbolic Interactionism Perspective,” Journal of Consumer Research, 319-329.

Tavassoli, N. T., Sorescu, A., and Chandy, R. (2014). Employee-based brand equity: Why firms with strong brands pay their executives less. Journal of Marketing Research, 51(6), 676-690.

Waters, L., McCabe, M., Kiellerup, D., and Kiellerup, S. (2002). The role of formal mentoring on business success and self-esteem in participants of a new business start-up program.

(36)

36

Appendix A: Rosenberg’s self-esteem scale

STATEMENT Strongly

Agree

Agree Neutral Disagree Strongly Disagree

1. I feel that I am a person of worth, at least on an equal plane with others.

☐ ☐ ☐ ☐ ☐

2. I feel that I have a number of good qualities.

☐ ☐ ☐ ☐ ☐

3. All in all, I am inclined to feel that I am a failure.

☐ ☐ ☐ ☐ ☐

4. I am able to do things as well as most other people.

☐ ☐ ☐ ☐ ☐

5. I feel I do not have much to be proud of.

☐ ☐ ☐ ☐ ☐

6. I take a positive attitude toward myself.

☐ ☐ ☐ ☐ ☐

7. On the whole, I am satisfied with myself.

☐ ☐ ☐ ☐ ☐

8. I wish I could have more respect for myself.

☐ ☐ ☐ ☐ ☐

9. I certainly feel useless at times. ☐ ☐ ☐ ☐ ☐

10. At times I think I am no good at all.

(37)

37

Appendix B: Survey

Thank you for participating in this survey.

This survey will consist of three parts and will only take 7 minutes of your time. Firstly some general questions will be asked after which you have to give your perferred choice regarding some job offers. Lastly some questions will be asked about you as a person. All information filled in will be handled with care and your response will be completely anonymous.

Daniel Hoogeboom. Masters student at the University of Groningen

/Page end/

What is your gender? - Male

- Female What is your age?

- Open textbox

What is your educational status - Student

- Working - Job seeking - None of the above

/Page end/

This second part presents you with a choice set for a job offer. Different elements of a job have been presented in varying levels. Please choose the job that is most suited to your preferences. If none of the options are interesting, you can pick "no-choice" which is the right most option

/Page end/

(38)

38

Choiceset 1

Salary: 2650 euro 2800 euro None-option

Possibilities for promotion and training:

There are possibilities for growth and training after 2 years

Contractual: It is a contract for 6

months It is a yearly contract

Company size and branding:

You will work in a medium-sized

organization with a locally known brand and multiple establishments in the region

You will work at a large multinational with a well-known brand and many establishments and employees around the world.

Choiceset 2

There are possibilities for growth and training after 1 year

Contractual: It is a two-year contract It is a yearly contract

You will work at a small unknown company with an unknown brand and only a few employees present

Choiceset 3

Contractual: It is a two-year contract It is a yearly contract

(39)

39

Choiceset 4

months It is a two-year contract

Choiceset 5

Contractual: It is a temporary contract It is a contract for 6 months

Choiceset 6

months It is a temporary contract

(40)

40

Choiceset 7

Contractual: It is a yearly contract It is a temporary contract

Choiceset 8

Contractual: It is a two-year contract It is a temporary contract

Choiceset 9

Contractual: It is a temporary contract It is a contract for 6 months

(41)

41

Choiceset 10

Contractual: It is a yearly contract It is a contract for 6 months

/Page end/

The survey will now list some general questions about you as a person. You have the option of choosing strongly agree if this really relates to you, agree if this relates in some way to you, disagree if this does not relate that much to you and strongly disagree if this does not relate to you at all.

- (See appendix A)

(42)

42

Appendix C: Utility levels for nominal model

Attribute Level Estimated utility value

Wald Pvalue Range Relative importance Salary 2500 -0.4926 33.9052 <0.0001 0.8368 26.13% 2650 -0.1322 2800 0.2806 2950 0.3442 Training 1 year 0.4853 47.3587 <0.0001 0.9181 28.67% 2 years -0.0526 3 years -0.4328 Contractual Part-time -0.4325 36.7662 <0.0001 0.8824 27.55% 6 months -0.2500 1 year 0.2327 2 years 0.4499 Equity Small -0.2414 19.2277 <0.0001 0.5654 17.65% Medium 0.3240 High -0.0827 None opt. 2.7278 12..2851 <0.0001 2.7278

Note: R-squared of this model is 0.3543

(43)

43

Appendix D: Hit rate box for all three models

The predictive validity of the model can be estimated in LatentGold. This too provides an internal prediction using 610 replications.

Table 1: Predictive statistics for the Linear Model

Observed/ Estimated 1 2 3 Total 1 190 98 0 288 2 87 217 0 304 3 9 9 0 18 Total 286 324 0 610

Note: Hit rate is calculated based on correctly estimated observations

Table 2: Predictive validity table Part-worth model

Note: Hit rate is calculated based on correctly estimated observations

Table 2: Predictive validity table moderated model

Measuring the impact of the personality trait self-esteem on job attributes and job choice

Measuring the impact of the personality

trait self-esteem on job attributes and job

choice

J.G.J.D. Hoogeboom

Measuring the impact of the personality

trait self-esteem on job attributes and job

choice

J.G.J.D. Hoogeboom

Management Summary

Contents

Abstract

1. Introduction

1.1 Job attributes

1.2 Personality

1.3 Marketing / managerial relevance

2. Theoretical background

2.1 Job search and choice

2.2 Personality; Self-esteem

3. Methodology

3.1 Sampling

3.2 Procedure and measures

3.3 Attributes and levels

3.4 Experimental design

3.5 Model

4. Results

4.1 Sample

4.2 Reliability

4.3 Main effects Model

4.4 Moderation effect

5. Conclusion and Discussion

References

Appendix A: Rosenberg’s self-esteem scale

Appendix B: Survey

Appendix C: Utility levels for nominal model

Appendix D: Hit rate box for all three models