Final Project: Test Report 1 Final Project: Test Report Diana L. Boyd Tests and Measurement PSYC-6315-1 Walden University Final Project: Test Report 2 Results Description of Construct When Donald Trump was elected President on November 9, 2016, he had the lowest minority support of any winning Presidential candidate in recent decades. He won with 28 percent of the Hispanic vote and only 8 percent of the African-American vote (Eisler, 2016). Strong opposition to any President or Presidential candidate is not unprecedented, but the high level of African-American opposition to Donald Trump has been remarkably steadfast. A Pew Research Center (2017) survey from February found Trump’s approval rating among blacks to be 14%. According to a Parker, Horowitz, & Mahl, (2016). Pew Research Center (2017) survey conducted in April, Trump’s approval rating among blacks was 14%. A June Pew Research Center (2017) survey revealed that Trump’s approval rating had sunk to 7%. Such a consistently high level of opposition among black Americans constitutes an alarming pattern of behavior that merits further study. In defining the construct, I considered the factors that could motivate African-American resistance towards Donald Trump and his administration. I researched the literature on voting patterns and political behavior to identify the major factors that influence political attitudes. Demographic characteristics—age, gender, income, educational status, religious preference, and region—are key variables in explaining political behavior in general, and thus, are incorporated into my model of the construct (Flanigan, Zingale, & Theiss-Morse, 2014). Besides demographic variables, attitudes towards political figures are also influenced by concerns about (1) the state of the economy, (2) views on domestic and foreign policy, and (3) personal characteristics of the President. Consistent with theories of political opinion formation, the construct of AfricanAmerican opposition to President Trump includes these three dimensions. Final Project: Test Report 3 The last element of the construct is racial anxiety. Group identity is an incredibly powerful force in politics (Wilson, 2012). Racial identity was undoubtedly the reason why African-Americans voted in such overwhelming numbers for Barack Obama, but the dynamics of identity politics are usually more complex (Block, 2011). In the case of Donald Trump, the climate of racial anxiety among African-Americans goes beyond mere identity politics. To many African-Americans, Trump’s election represented what has been called a “white-lash” – a backlash against black Americans and the policies of the nation’s first African-American President (Madhani, 2016). Since the election, there has been a well-documented rise in raciallymotivated violence in America (Okeowo, 2017). Many trace these incidents of racist violence back to Trump and his inflammatory rhetoric (Chen, 2017). Any instrument that measures this construct must consider the effect of racial anxiety on African-American opposition. Test Specifications Development of the African-American Opposition to Trump Scale (AAOTS) proceeded along two lines. The initial task was to identify factors that affect political behavior regardless of race. After researching political behavior and voting patterns in U. S. Presidential Elections, I identified key demographic traits that have been known to shape voting behavior. These factors include age, gender, income, educational status, religious preference, marital status, and region. These social factors are measured in Item 1. The remaining items represent the heart of the AAOTS. These items measure attitudes related to Donald Trump’s personal characteristics, stances on issues that define Trump’s agenda, concerns about the state of the economy, and attitudes on racial issues. The American National Election Studies (2016) questionnaires served as a model for writing items on these topics. The second line of development concentrated on the impact of racial anxiety on African-Americans. Items were modeled after items from a Pew Final Project: Test Report 4 Research Center (2016) survey from June on race relations in America (Parker, Horowitz, & Mahl,2016). The original questionnaire contained 24 items. Before the questionnaire was finalized, it was given to a group of seven doctoral students enrolled in the Psychology and Counseling program at Walden University. Additionally, five faculty members at several universities with expertise in the area of race, political behavior, and scale construction were consulted to verify that test items accurately reflected the content domain. Items that were approved by a majority of the experts as congruent with the content domain were retained. As a result, several items were reworded for clarity and parsimony. The first item on the test compiles demographic information about the participant. The actual AAOTS consists of 20 items rated on a 5-point Likert scale with four subscales of 5 items each - Trump’s Personality, Policy Orientation, Economic Perceptions, and Racial Anxiety. These subscales correspond to the four factors in my model of African-American opposition to Trump. The test will be administered to a random sample of 385 African-Americans in the 50 states plus District of Columbia. Participants will be selected through stratified random-digit dialing (RDD) using standard list-assisted methods. Interviewers will introduce the project to the youngest adult living at the selected household. If that individual expresses interest, he or she will be screened for eligibility to determine whether they self-identify as African American and are age 18 or older. This process is carried out until the sample mirrors the population in terms of age and gender. The initial screening consists of several generic personal political questions including whether or not they approve or disapprove of Donald Trump’s handling of the Presidency. Respondents who disapprove of Donald Trump will be contacted through mail and Final Project: Test Report 5 incentivized to participate in a face-to-face interview at which time the AAOTS questionnaire will be given. Test Items African-American Opposition to President Trump Scale Test Format: This scale consists of 21 items that assess the nature of opposition among African-Americans who disapprove of the way President Trump is handling his job as President. Item 1 is used to gather demographic information about the participant. The scale proper starts at item 2. This test operates on the assumption that opposition to the President consists of four factors – personal qualities of President, the administration’s major policies and positions, economic perceptions, and racial anxiety. Each of these factors is measured by a separate subscale – Trump’s Personality (#2-6), Policy Orientations (#7-11), Economic Perceptions (#1216), and Racial Anxiety (#17-21). Demographic information: 1. Age of respondent: Gender: Family Income: Educational Level: Religion: Region: How well do the following phrases describe your impressions of President Trump so far? 2. He is trustworthy 3. He is well-informed 4. He is able to get things done Final Project: Test Report 5. He is a good communicator 6. He respects the country’s institutions What is your level of support or opposition to the following policies? 7. Participation in Paris climate agreement to limit global warming 8. Tax cuts across all income levels plus businesses 9. Building a wall across the entire border with Mexico 10. The Affordable Care Act of 2010 (Healthcare act passed by Obama in 2010) 11. A temporary ban on travelers to the United States from Muslim-majority nation How much do you agree or disagree with the following statements? 12. I am better off financially than I was a year ago 13. A year from now I expect to be better off financially 14. The economy is in a better state than it was a year ago 15. A year from now I expect the economy to be in a better state 16. There are more job opportunities than there were a year ago How much do you agree or disagree with the following statements? 17. Race relations in America are good 18. Race relations are getting better 19. Blacks and whites have equal rights and opportunities 20. Black people are treated fairly by the police 21. Racially-motivated violence has not increased in the last year Descriptive Statistics 6 Final Project: Test Report 7 Descriptive statistics for the main variables in the study, including means and standard deviations, are reported in Table 1. Of the 1,146 participants in the Money Data survey, 356 were male and 790 were female. 54.6% of participants were married at the time of the survey and 45.4% were single. Participants’ mean income was $44,247 with a standard deviation of $52,992. The mean age of participants was 32.9. Of note, the mean value for item L6R (“I'd rather have a modest lifestyle because it's less stressful”), which is the recoded version of L6, and item D5 (“I expect others to help me out when I have a financial problem”) are lower than the other items. Item R6 (“Running a business is something that I think of as interesting and exciting”) had a rather high mean value compared to other items on questionnaire. Factor Analysis Factor analysis is a statistical technique in which many variables are represented by a few clusters of variables called factors (Darlington, 2004). On a practical level, factor analysis makes a complex set of data simpler and more manageable. Confirmatory factor analysis, which is beyond the scope of this project, aims to compare the researcher’s theoretical model with the observed data (Yong & Pearce, 2013). In exploratory factor analysis, the goal is simply to discover the factors that account for the correlations within the data set. Exploratory factor analysis will be the method we use to analyze the Money Data scale. I carried out an exploratory factor analysis in SPSS using the Principal Axis extraction method. I selected “Direct Oblimin” because I expected some of the factors to be correlated with one another. I selected “Fixed number of factors” and entered “3” because I expected that there would be three factors corresponding to the three subscales on the Money Data questionnaire. Final Project: Test Report 8 Table 2A shows that all the assumptions of the factor analysis have been met. The determinant of the correlation matrix is .009, well above the cut-off of .0001. This shows that multicollinearity is not a problem for the data and there is no need to eliminate any of the variables. The KMO Measure of Sampling Adequacy is 0.750, which is greater than the threshold of 0.70. Bartlett’s Test of Sphericity checks whether the correlation matrix is significantly different than the identity matrix, in which case there would be zero correlation between different variables (Leech, Barrett, & Morgan, 2005). The p value is <.001 and we can reject the null hypothesis that the correlation matrix is the identity matrix. Table 2B contains information about the amount of variance accounted for by each factor. Out of 18 possible factors, just 6 have eigenvalues greater than 1.0. Initial eigenvalues associated with each factor show the approximate amount of variance accounted for in terms of the “item’s worth” of variance each factor explains (Leech, Barrett, & Morgan, 2005). Factor 3, for example, has an initial eigenvalue of 1.985 and accounts for two items’ worth of variance. The percentage of variance accounted for by each factor is also shown in this table. Factor 1 accounts for 19.935% of the total variance, Factor 2 accounts for 13.620%, Factor 3 accounts for 11.030%, Factor 4 accounts for 7.016%, Factor 5 accounts for 6.246%, and Factor 6 accounts for 5.827%. Those 6 factors account for a total of 63.674% of the variance. The three factors I have extracted—Factor 1, Factor 2, and Factor 3—account for 44.585% of the variance. The sums of square loadings after rotation are 2.671, 1.711, and 2.265 respectively. The scree plot, displayed in Table 2C, is a graphical representation of the distribution of variance among the factors. The location of the “elbow” of the curve is the point at which subsequent factors contribute diminishing additional variance. There is a marked decrease in Final Project: Test Report 9 slope between the third and fourth factor. We could have justified a four- factor solution, but for reasons of simplicity the three-factor solution is preferred. The pattern matrix shown in Table 2D is the key to interpreting the factor analysis. The pattern matrix shows that each item loads onto one of the three factors without any significant cross-loadings. We can see from the Table 2D that items L1, L2, L3, L4, L5, and L6 are loaded onto Factor 3. L3 has a smaller loading compared to the other items (-.342), but is within the cutoff range of 0.3-0.4. Items D1, D2, D3, D4, D5, and D6 are loaded onto Factor 2. D4 has a low factor loading, but is above the minimum threshold. Items R1, R2, R3, R4, R5, and R6 are all loaded onto Factor 1. R2 and R3 have quite high factor loadings. R5 is has a low factor loading (.315), but is above the cut-off point. Naming factors is more art than science, but attaching labels to the individual factors allows us to assess face and construct validity, and hence, is an essential step in the process (Yong & Pearce, 2013). Items R1-R6 all load onto Factor 1 and these items measure financial Risk-Taking or financial aggressiveness in one way or another. I will label this factor “RiskTaking” for short. Items D1-D6 all load onto Factor 2 and this factor will be referred to as “Dependency” because these items measure degrees of financial Dependency. Items L1-L6, which all concern the desire for luxury or privilege, load onto Factor 3. I will therefore label this factor “Lifestyle”. Each factor represents a subscale on the test. From this point on, each factor will be referred to by its label. Because I selected an oblique rotation (“Direct Oblimin”), the output generates a “Factor Correlation Matrix”, which is included in Table 2E. The “Factor Correlation Matrix” shows how each factor correlates with the other two factors. Roughly speaking, it quantifies how much “overlap” there is between factors (Yong & Pearce, 2013). If two factors are highly correlated, it Final Project: Test Report 10 might be possible to combine them into a single factor. I expected that there some of the factor are correlated, but not to a large degree. None of the factors are highly correlated. The largest correlation is between Factor 1 and Factor 3 (-.293). In conclusion, factor analysis indicates that the items compromising the psychological test “Money Data” load onto three distinct factors. Items R1, R2, R3, R4, R5, and R6 all load onto the Risk-Taking factor. Items D1, D2, D3, D4, D5, and D6 all load onto the Dependency factor. Items L1, L2, L3, L4, L5, and L6 all load onto the Lifestyle factor. Despite some overlap between the Factor 1 and 3, the three factors are distinct. Reliability In psychometrics, reliability is “degree to which test scores are free from error and yield consistent results” (Thanasegaran, 2009). The basic theory of test reliability posits that observed test scores depend on two factors – the individual’s “true score” and errors of measurement. An individual’s true score is the characteristic that the researcher is trying to measure whereas errors of measurement contribute to inconsistency in the observed test scores (Murphy & Davidshofer, 1988). A reliability coefficient measures the ratio of true score’s variance to the total variance of the test scores. While the definition of the reliability coefficient is straightforward, its interpretation is not. Several methods have been developed to estimate reliability but can be grouped into four general categories – test-retest methods, alternate form methods, split-half methods, and internal consistency methods. In estimating the reliability of the Money Data questionnaire, I will be primarily concerned with internal consistency methods. Internal consistency methods measure how well the items on a test measure the construct under consideration (Hale & Astolfi, 2011). The most popular of these methods is Cronbach’s coefficient alpha. This is the index of reliability I will use to assess the reliability of the Money Final Project: Test Report 11 Data test. Given that we have identified three subscales in our factor analysis, I will calculate Cronbach’s alpha for the three subscales – Risk-Taking, Dependency, and Lifestyle – and then for the test as a whole. Cronbach’s alpha is 0.786 for the Risk-Taking scale. Coefficients with values greater than 0.7 are considered strong so this result is very satisfactory (Thanasegaran, 2009). Table 3A shows the value of Cronbach’s alpha if each item was deleted from the scale. We immediately notice that the value of Cronbach’s alpha would decrease if items R1, R2, R3, R4, and R6 were deleted, however, Cronbach’s alpha would increase if R5 were deleted. Recall from the factor analysis section (see above) that R5 had a small loading. I would recommend that this item is deleted from the questionnaire because it does not seem to be measuring the same construct as the other items on the scale. For the Lifestyle scale, Cronbach’s alpha is 0.732, which is satisfactory. Table 3B shows that the reliability coefficient for this scale decreases if L1, L2, L4, L5, or L6 are deleted, but increases to 0.745 if L3 is deleted. Because this item’s deletion does not represent a drastic change in the coefficient alpha, it might just need to be rewritten. Compound items are discouraged on questionnaires plus the inclusion of the phrase “higher quality” might be problematic by itself. Moreover, people tend to think of the products they purchase as “higher quality” regardless of the price. I recommend revising the item to “I tend to buy more expensive products”. The Cronbach’s alpha is 0.680 for the Dependency scale. There is some disagreement over what is the lower limit for an acceptable alpha coefficient. Some authors consider a coefficient above 0.7 to be acceptable while others recommend a lower limit of 0.65 (Thanasegaran, 2009; see also Goforth, 2015). Table 3C shows that the reliability decreases if Final Project: Test Report 12 any of the items are deleted. There are no weak items on the Dependency scale. Although the reliability of the Dependency scale is not unacceptably low, it could be improved. I would consider adding more items to the scale as test length tends to improve reliability (Cortina, 1993). Cronbach’s alpha for the entire test was found to be 0.723. This is an acceptable level of reliability. Validity Construct validity of the Money Data scale was assessed through exploratory factor analysis (see above for more details). Three factors were retained – the Risk-Taking factor, the Dependency factor, and the Lifestyle factor. Together they account for 44.59% of the variance in the total variance. These three dimensions correspond to the three subscales of the test. The subscales were significantly correlated to one another as expected, but not to the level where they would need to be merged. Concurrent validity was tested by examining the Pearson correlation coefficients between the three factors retained from the factor analysis and the Financial Comfort and Relationship Happiness subscales. This was accomplished by summing the raw scores corresponding to all items loading on a factor and then testing the correlation coefficient for statistical significance. Predictive validity was tested by examining the correlations between each scale and income. The correlation matrix is shown in Table 4A. I hypothesized that the Risk-Taking factor was negatively correlated with Relationship Happiness. The Risk-Taking factor was tested for statistical significance and found to be negatively correlated with Financial Happiness (r = -.091, p < .01). The level of correlation is very weak. Final Project: Test Report 13 I hypothesized that the Risk-Taking factor was positively correlated with the participant’s income. A test of statistical significance showed that there was a weak positive correlation (r = .131, p < .01). I hypothesized that Dependency would be negatively correlated with Financial Comfort. A test of statistical significance showed that there was a weak positive correlation (r =- .173, p < .01). I anticipated that the Dependency factor would have a complicated relationship to income. The Dependency should decrease as a participant’s income increases. On the other hand, the magnitude of the difference between the partner’s income and a participant’s income would seem to be the crucial factor. Because increases in the utility of income are nonlinear, measuring income disparities between partners can be quite complicated. For example, a difference of $20,000 in a relationship when the combined income is $30,000 is much more severe than a disparity of $20,000 when the combined income is $150,000. To avoid these complications, I simply examined the correlations between Dependency and the participant’s income separately from Dependency and the partner’s income. As stated earlier, as a participant’s income increases, Dependency should decrease. This is born out statistically as the two variables are weakly negatively correlated (r = -.192, p < .01). I hypothesized that the relationship between a partner’s income and Dependency should have a positive linear relationship. This relationship was verified (r = .120, p < .01). Although this method is somewhat simplistic, the result is what we would expect. I hypothesized that the Lifestyle factor was positively correlated with Financial Comfort. Naturally, the more comfortable one is financially, the greater the desire for Lifestyle. The relationship between the Lifestyle factor and Financial Comfort was not statistically significant, Final Project: Test Report though (r = .053, p =.073). This might be an anomaly because Lifestyle and a participant’s income are significantly correlated (r = -0.297, p < .01). 14 Final Project: Test Report 15 References American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. Washington, DC: American Educational Research Association. American National Election Studies (www.electionstudies.org). 2010. Time Series Cumulative Data File [dataset]. Stanford University and the University of Michigan [producers and distributors]. Block, R. (2011). Backing Barack because he's Black: Racially motivated voting in the 2008 election. Social Science Quarterly, 92(2), 423-446. Charity, J. (2016, November 10). Why Donald Trump Terrifies Many Black Americans – The Ringer. Retrieved July 02, 2017, from https://theringer.com/why-donald-trump-terrifiesmany-black-americans-c3cf008c78e6 Chen, M. (2017, March 24). Donald Trump's Rise Has Coincided With an Explosion of Hate Groups. Retrieved August 03, 2017, from https://www.thenation.com/article/donaldtrumps-rise-has-coincided-with-an-explosion-of-hate-groups/ Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of applied psychology, 78(1), 98. Darlington, R. B. (2004). Factor Analysis. Retrieved June 19, 2017, from http://node101.psych.cornell.edu/Darlington/factor.htm Eisler, P. (2016, November 23). Trump won with lowest minority vote in decades, fueling divisions. Reuters. Retrieved from http://reuters.com Flanigan, W. H., Zingale, N. H., Theiss-Morse, E. A., & Wagner, M. W. (2014). Political behavior of the American electorate. Cq Press. Final Project: Test Report 16 Goforth, C. (2015, November 16). Using and Interpreting Cronbach’s Alpha. Retrieved July 22, 2017, from http://data.library.virginia.edu/using-and-interpreting-cronbachs-alpha/ Hale, C. D., & Astolfi, D. (2011). Measuring learning and performance: A primer. Leech, N. L., Barrett, K. C., & Morgan, G. A. (2005). SPSS for intermediate statistics: Use and interpretation. Psychology Press. Lester, P. E., & Bishop, L. K. (2000). Handbook of tests and measurement in education and the social sciences. Lanham, Md: Scarecrow Press. Madhani, A. (2016, November 10). Trump's victory leaves black community reeling. Retrieved July 25, 2017, from https://www.usatoday.com/ Morgan, G. A., Leech, N. L., Gloeckner, G. W., & Barrett, K. C. (2012). IBM SPSS for introductory statistics: Use and interpretation. Routledge. Murphy, K. R., & Davidshofer, C. O. (1988). Psychological testing. Principles, and Applications, Englewood Cliffs. Parker, K., Horowitz, J., & Mahl, B. (2016). On Views of Race and Inequality, Blacks and Whites Are Worlds Apart Washington, DC: Author. Retrieved from http://assets.pewresearch.org/wp-content/uploads/sites/5/2017/04/17171737/04-17-17Political-release.pdf Pew Research Center. (2017). In the first months, views of Trump are strongly felt, deeply polarized Washington, DC: Author. Retrieved from http://assets.pewresearch.org/wpcontent/uploads/sites/5/2017/02/17094915/02-16-17-Political-release.pdf Final Project: Test Report 17 Pew Research Center. (2017). Public has criticism of both parties, but Democrats lead on empathy for middle class.Washington, DC: Author. Retrieved from http://assets.pewresearch.org/wp-content/uploads/sites/5/2017/06/20110029/06-20-17Political-release-1.pdf Public Dissatisfaction With Washington Weighs on the GOP. (n.d.). Retrieved August 06, 2017, from http://latinosreadytovote.com/public-dissatisfaction-washington-weighs-gop/ Thanasegaran, G. (2009). Reliability and Validity Issues in Research. Integration & Dissemination, 4. Wilson, J. M. (2012). How are we doing? Group-based economic assessments and African American political behavior. Electoral Studies, 31(3), 550-561. Yong, A. G., & Pearce, S. (2013). A beginner’s guide to factor analysis: Focusing on exploratory factor analysis. Tutorials in quantitative methods for psychology, 9(2), 79-94. Final Project: Test Report 18 Appendix Descriptive Statistics Table 1 Std. Minimum Maximum Mean Deviation Participant's income $0 $500,000 $44,247.41 $52,992.329 Partner's income $0 $400,000 $45,486.10 $45,414.748 Financial Comfort 1.00 2.00 1.5419 .49846 Participant's age 18 68 32.90 11.250 Partner's age 18 69 34.11 11.402 Participant's gender 1.00 2.00 1.6894 .46296 Partner's gender 1.00 2.00 1.3298 .47036 Participant's 1.00 7.00 3.8525 1.33939 Partner's education 1.00 7.00 3.5151 1.43283 I need luxury to feel 1.00 5.00 2.4433 1.04016 1.00 5.00 2.4625 1.08136 1.00 5.00 2.4930 1.14474 1.00 5.00 2.7818 1.16329 1.00 5.00 2.6073 1.16847 1.00 5.00 4.0436 1.12138 1.00 5.00 1.9564 1.12138 education comfortable. I enjoy having people wait on me. I tend to buy more expensive, higher quality products. I like to give other people the impression that I am financially well-off. To tell the truth, I like people to envy my financial status. I'd rather have a modest lifestyle because it's less stressful. L6R Final Project: Test Report My partner tends to 19 1.00 5.00 2.7688 1.33535 1.00 5.00 2.3019 1.25991 1.00 5.00 2.1396 1.22052 1.00 5.00 2.3185 1.21732 1.00 5.00 1.8621 1.03907 1.00 5.00 2.8578 1.25584 1.00 5.00 3.1632 1.35508 1.00 5.00 2.4538 1.17843 1.00 5.00 2.7827 1.19858 1.00 5.00 2.4258 1.14235 be more responsible with money than I am. I've never been very good at handling financial responsibility I expect my partner to take care of me financially. I know that my partner has more financial responsibility than I do, so I try not to "rock the boat." I expect others to help me out when I have a financial problem. I'm more of a nurturer than a provider. I'd rather run my own business than to work for someone else. I don't mind risking large amounts of money if there is a good chance I can come out ahead. I'm willing to take real chances to get ahead financially. I get bored unless I'm taking some risks with my career. Final Project: Test Report Being too 20 1.00 5.00 2.8307 1.04533 1.00 5.00 3.4284 1.28295 Marital Status .00 1.00 .5462 .49807 Relationship 1.00 5.00 3.9101 1.05968 LIFE 6.00 30.00 14.7443 4.39671 RISKTAKE 6.00 30.00 17.0846 5.02299 DEPENDENT 6.00 30.00 14.2487 4.55707 conservative with your investments can cause financial problems. Running a business is something that I think of as interesting and exciting. Happiness Factor Analysis Table 2A KMO and Bartlett's Test Kaiser-Meyer-Olkin Measure of .750 Sampling Adequacy. Bartlett's Test of Approx. Chi- Sphericity Square 5335.789 df 153 Sig. .000 Table 2B Total Variance Explained Rotation Sums of Squared Loadingsa Initial Eigenvalues Factor Total % of Variance Cumulative % Total 1 3.588 19.935 19.935 2.671 2 2.452 13.620 33.555 1.711 3 1.985 11.030 44.585 2.265 4 1.263 7.016 51.601 5 1.124 6.246 57.847 Final Project: Test Report 21 6 1.049 5.827 63.674 7 .864 4.800 68.475 8 .779 4.328 72.803 9 .716 3.977 76.780 10 .633 3.517 80.297 11 .601 3.337 83.635 12 .550 3.053 86.687 13 .546 3.031 89.719 14 .480 2.667 92.386 15 .402 2.232 94.618 16 .348 1.932 96.550 17 .331 1.837 98.388 18 .290 1.612 100.000 Table 2C Final Project: Test Report 22 Table 2D Pattern Matrixa Factor 1 I need luxury to 2 3 -.016 .082 -.648 .028 -.097 -.648 -.048 .080 -.342 .009 -.117 -.615 .047 .025 -.581 .064 .030 -.524 .060 .425 -.016 .080 .449 .023 -.115 .590 -.062 feel comfortable. I enjoy having people wait on me. I tend to buy more expensive, higher quality products. I like to give other people the impression that I am financially well-off. To tell the truth, I like people to envy my financial status. I'd rather have a modest lifestyle because it's less stressful (recoded) My partner tends to be more responsible with money than I am. I've never been very good at handling financial responsibility I expect my partner to take care of me financially. Final Project: Test Report I know that my 23 .024 .641 .083 .039 .367 -.104 -.138 .592 .024 .648 .069 .050 .750 -.046 -.056 .763 -.002 -.033 .585 .037 -.042 .315 -.102 -.058 .620 .075 .068 partner has more financial responsibility than I do, so I try not to "rock the boat." I expect others to help me out when I have a financial problem. I'm more of a nurturer than a provider. I'd rather run my own business than to work for someone else. I don't mind risking large amounts of money if there is a good chance I can come out ahead. I'm willing to take real chances to get ahead financially. I get bored unless I'm taking some risks with my career. Being too conservative with your investments can cause financial problems. Running a business is something that I think of as interesting and exciting. Final Project: Test Report 24 Extraction Method: Principal Axis Factoring. Rotation Method: Oblimin with Kaiser Normalization.a a. Rotation converged in 5 iterations. Table 2E Factor Correlation Matrix Factor Risk-Taking Dependence Lifestyle Risk-Taking 1.000 -.016 -.293 Dependence -.016 1.000 -.122 Lifestyle -.293 -.122 1.000 Reliability Table 3A Item-Total Statistics I'd rather run my Scale Mean if Item Scale Variance if Corrected Item-Total Squared Multiple Cronbach's Alpha if Deleted Item Deleted Correlation Correlation Item Deleted 13.9215 17.065 .565 .473 .747 14.6309 17.410 .654 .515 .725 14.3019 17.196 .664 .535 .722 14.6588 18.768 .521 .301 .757 own business than to work for someone else. I don't mind risking large amounts of money if there is a good chance I can come out ahead. I'm willing to take real chances to get ahead financially. I get bored unless I'm taking some risks with my career. Final Project: Test Report Being too 25 14.2539 21.383 .285 .129 .805 13.6562 17.805 .534 .450 .754 conservative with your investments can cause financial problems. Running a business is something that I think of as interesting and exciting. Table 3B Item-Total Statistics – Lifestyle Scale I need luxury to Scale Mean if Item Scale Variance if Corrected Item-Total Squared Multiple Cronbach's Alpha if Deleted Item Deleted Correlation Correlation Item Deleted 12.3010 14.083 .534 .299 .677 12.2818 13.946 .522 .449 .679 12.2513 15.428 .288 .091 .745 11.9625 13.849 .477 .422 .692 12.1370 13.474 .524 .401 .677 12.7880 14.082 .474 .379 .693 feel comfortable. I enjoy having people wait on me. I tend to buy more expensive, higher quality products. I like to give other people the impression that I am financially well-off. To tell the truth, I like people to envy my financial status. I'd rather have a modest lifestyle because it's less stressful (recoded) Table 3C Final Project: Test Report 26 Item-Total Statistics Scale Mean if Item Scale Variance if Corrected Item-Total Squared Multiple Cronbach's Alpha if Deleted Item Deleted Correlation Correlation Item Deleted I'd rather run my 13.9215 17.065 .565 .473 .747 14.6309 17.410 .654 .515 .725 14.3019 17.196 .664 .535 .722 14.6588 18.768 .521 .301 .757 14.2539 21.383 .285 .129 .805 13.6562 17.805 .534 .450 .754 own business than to work for someone else. I don't mind risking large amounts of money if there is a good chance I can come out ahead. I'm willing to take real chances to get ahead financially. I get bored unless I'm taking some risks with my career. Being too conservative with your investments can cause financial problems. Running a business is something that I think of as interesting and exciting. Table 4A Correlations Participant's income Pearson Participant's Partner's income income 1 .344 Financial Relationship Comfort ** .164 happiness ** risk depend lifestyle ** -.192** .138** .000 .000 .000 .043 .131 Correlation Sig. (2-tailed) .000 .000 .148 Final Project: Test Report N Partner's income Pearson 27 1146 1146 1146 1146 1146 1146 1146 .344** 1 .162** .016 .031 .120** .029 .000 .595 .302 .000 .326 1146 1146 1146 .009 -.173** .053 .000 .764 .000 .073 1146 1146 1146 - -.009 .002 .002 .753 .941 Correlation Financial Comfort Sig. (2-tailed) .000 N 1146 1146 1146 1146 .164** .162** 1 .216** Sig. (2-tailed) .000 .000 N 1146 1146 1146 1146 .016 ** 1 Pearson Correlation Relationship happiness risk Pearson .043 .216 Correlation .091 ** Sig. (2-tailed) .148 .595 .000 N 1146 1146 1146 1146 1146 1146 1146 ** 1 -.011 .254** .700 .000 ** .031 .009 Sig. (2-tailed) .000 .302 .764 .002 N 1146 1146 1146 1146 1146 1146 1146 -.192** .120** -.173** -.009 -.011 1 .087** Sig. (2-tailed) .000 .000 .000 .753 .700 N 1146 1146 1146 1146 1146 1146 1146 .138** .029 .053 .002 .254** .087** 1 Sig. (2-tailed) .000 .326 .073 .941 .000 .003 N 1146 1146 1146 1146 1146 1146 Pearson .131 -.091 Correlation depend Pearson Correlation lifestyle Pearson .003 Correlation 1146