Research on Graduate Attributes in Education Psychology

Prof. Gavin T. L. Brown The University of Auckland Correspondence: gt.brown@Auckland.ac.nz Presentation to the 10thEducational Psychology Forum, Wellington, November 2017 Available: https://doi.org/10.17608/k6.auckland.5630602

UoA Vice-Chancellor Strategic Development Fund which supported Dr Makayla Grays as Post-Doctoral Research Fellow. Makayla did all the data collection and was a key player in the design of the studies and data instruments and analyses. Dr Grays is now at Connecticut College, USA.

1stdegree students will have a range of skills, attitudes, dispositions, and knowledge upon graduation In Australasia called graduate attributes University of Auckland Graduate Profile 2003-2016 I Specialist II General intellectual skills and III Personal qualities Specific focus of this research 2.2 3.1 3.5 an appreciation of human and cultural diversity; I Specialist knowledge II General intellectual skills and capacities III Personal qualities knowledge capacities 2.2 3.1 3.5 an appreciation of human and cultural diversity; intellectual openness and curiosity; Love and enjoyment of ideas, discovery and learning Respect for the values of other individuals and groups, and intellectual openness and curiosity; Love and enjoyment of ideas, discovery and learning; ; Respect for the values of other individuals and groups, and

Attributes 2.2 + 3.1 Standardised pattern coefficients and error variances of the 15-item, three- factor model in Late 2014; Early 2014 estimates superscripted

Attribute 3.5 Standardised pattern coefficients and error variances of the 15-item, three- factor model in Late 2014; Early 2014 estimates superscripted

Graduate Diploma students in teacher education compared to undergraduate students in initial teacher education had stat. sig. and practically larger means Love of learning (d=.36) Openness to diverse cultures/groups (d=.80) But is it because they were older? Already had a degree? Personality? Not sure

Re-administer questionnaire with new cohort of ITE students Use personality as a predictor of graduate attributes Expect that Openness and Conscientiousness especially would predict greater curiosity and openness to diversity Need to use a brief inventory to avoid fatigue, abandonment, missing data

Openness imagination and insight, and those high in this trait also tend to have a broad range of interests Extraversion excitability, sociability, talkativeness, assertiveness, and high amounts of emotional expressiveness Agreeableness trust, altruism, kindness, affection, and other prosocial behaviors; tend to be more cooperative Conscientiousness high levels of thoughtfulness, with good impulse control and goal-directed behaviors Neuroticism sadness, moodiness, and emotional instability; tend to experience mood swings, anxiety, irritability and sadness Purportedly universal human traits Norman, 1963 Openness Extraversion Agreeableness Conscientiousness Neuroticism

A brief 10-item inventory (Rammstedt, & John, 2007). two items per trait, one of which is reverse scored. Based on US and German samples, the BFI10 had good correlation with the 44 item version and relatively low factor inter-correlations (mean r=.11) supporting the validity and discrimination of the brief version.

Older analytic procedures Cronbach alpha estimates of scale reliability, exploratory factor analysis with varimax rotation, and inspection of scree plots modern standards for scale or inventory validation. Low reliance on alpha as an indicator of scale quality (Sijtsma, 2009), the use of scree plots for dimensionality detection (Courtney, 2013), use oblique rotation in EFA (Costello & Osborne, 2005) and confirmatory factor analysis (Hoyle & Duvall, 2004)

87% were women. 42% P keha, 19% were Asian, 12% Pasifika, and 20% gave more than one ethnicity. Year 1 = 66%; nearly 28% final year of undergraduate study Just over 80% were in ITE; 10% human services or social work. 58% aged under 25, and 14% aged 40 or over. Sample seems consistent with high proportion direct entry from high school into primary teacher education, and older group reflects tendency for mature students to return to higher education for either a career change or qualifications upgrade.

Online LimeSurvey: during the first 6 weeks of semester 1 of the academic year. Speed of responding used to remove 32 Questionnaires completed <3 seconds per item 298 valid cases with acceptable completion time of just under 5 minutes Response Scale 6 item positively-packed agreement (4 positive, 2 negative options) Positive: slightly, moderately, mostly, strongly; Negative: mostly, strongly Confirmatory factor analyses Maximum likelihood estimation because response scale had >5 options (Finney & DiStefano, 2006). Fit imputed if the ratio of 2 to df was statistically not significant, gamma hat >.90, and SRMR <.06 (Fan & Sivo, 2007; Marsh, Hau, & Wen, 2004). Items with weak loadings (i.e., <.30) were removed in graduate attribute factors. Structural equation modeling used to identify statistically significant (i.e., >.11) regressions from the BFI10 factors to the graduate attribute factors.

.89 Love of Learning I love to learn. I find learning to be interesting and exciting. A9 A4 A1 2 A1 A6 A1 5 .86 .83 .38 .68 Perspectives I enjoy being introduced to new ideas and ways of thinking. I try to do things that will broaden my perspective. B5 B9 B13 I have discussions with people about their ideas. I like to consider how other people's perspectives might differ from mine. Cultures B6 I want to know more about another ethnic group. I am interested in visiting sites that have special significance in another culture. B10 I read about customs and cultural practices of other groups. B14 I enjoy going places where people speak a different language. Backgrounds I like getting to know people of a race or nationality other than my own. B8 I enjoy conversations with people about their backgrounds. B16 I like thinking about how people are shaped by their experiences. B4 I attempt to learn about people's upbringing and life experiences. .83 .73 .68 .67 The prospect of learning new things excites me. I look forward to learning new things. I want to know more about things. .78 .74 .61 .67 B1 Learning more in my field of study pleases me. Curiosity I am curious about things. I am inquisitive. .91 .97 A8 A5 A1 1 A3 A1 4 .84 .77 .75 .72 B2 .75 .66 .59 I am intrigued by many different topics. Many things interest me. .71 .99 .68 I enjoy thinking about things. Answer Seeking .77 B12 .95 A1 0 A1 3 A2 A7 .76 .69 .52 .71 I like finding answers to questions. .68 I like to enquire about things I do not understand. I enjoy searching for answers. I like trying to solve problems that puzzle me. .68 .62 2 2/df (p) CFI Gamma Model k RMSEA 90%CI SRMR df hat .89 Marginal to good fit Attributes 27 809.29 317 2.55 (.11) .89 .072 .066-.078 .064

2 2/df (p) Model k CFI Gamma RMSEA 90%CI SRMR df hat .90 Joint 37 1210.10 607 1.99 (.16) .88 .058 .053-.063 .063 Acceptable to Good fit

Personality appears to be a statistically significant but small predictor of intellectual curiosity and openness to diversity. OPENNESS to new experiences contributed sensibly to intellectual curiosity seeking new knowledge, ideas, answers to problems, and a love of learning seem predicated on a desire to be open to new thinking. Similarly, in programs that have accountability evaluations (i.e., formative and summative course assessments and terminal professional certification), being CONSCIENTIOUS and AGREEABLE are favourable and self-regulating responses to these requirements. Seeking answers specifically requires tolerance for uncertainty and so the negative path from NEUROTICISM seems appropriate.

Negative error variances observed & fixed Marginal to good fit 2=101.46, df =29, 2/df =3.50 (p=.06), CFI=.76, Gamma hat=.95, RMSEA=.092 (90%CI=.073-.111), SRMR=.072, AIC=153.46 extremely weak item loadings (i.e., <.30) for three of the 10 items. scale alpha values were poor, especially for Openness The scale inter-correlations were generally weak (mean |r|=.21), though stronger than reported by Rammstedt and John (2007).

Extraversion rC1 (R) I see myself as someone who is reserved C6 I see myself as someone who is outgoing, sociable. Agreeableness rC7 (R) I see myself as someone who tends to find fault with others C2 I see myself as someone who is generally trusting. Conscientiousness rC3 (R) I see myself as someone who tends to be lazy C8 I see myself as someone who does a thorough job. Neuroticism C9 I see myself as someone who gets nervous easily. rC4 (R) I see myself as someone who is relaxed, handles stress well Openness C10 I see myself as someone who has an active imagination. rC5 (R) I see myself as someone who has few artistic interests .64 .63 .99 .21 .87 .36 .99 .28 .80 .10

Factor I. Extroversion II. Agreeableness III. Conscientiousness .38 IV. Neuroticism V. Openness Inter-correlations II. III. .01 (.35) .30 (.48) -.44 -.14 -.38 (.44) .26 .01 I. IV. V. M 3.92 1.09 4.67 0.91 4.46 1.00 2.37 1.09 SD (.58) .12 -.03 (.13) 4.26 0.99 Note. (.xx)=estimate of scale reliability

Rammstedt, Goldberg, and Borg (2010) recommended ipsative item responses Centering on own mean and adjusting by how different each factor was to overall individual mean NOT not definite covariance matrices High inter-correlations suggest fewer factors needed NOT admissible because of positive

Fit indices ranged from marginal to good ( 2=95.80, df =29, 2/df =3.30 (p=.07), CFI=.78, Gamma hat=.96, RMSEA=.088 (90%CI=.068-.108), SRMR=.068, AIC=147.80). pairs of items did not match the original Model 1. 3 items rotated around the big five factors: Agreeable gained item 5 (I see myself as someone who has few artistic interests) from Openness item 7 (I see myself as someone who tends to find fault with others) went to Conscientiousness, item 8 (I see myself as someone who does a thorough job) went to Openness

modification indices indicated Openness factor should be removed, Deletion of items 8 (I see myself as someone who does a thorough job) and 10 (I see myself as someone who has an active imagination). marginal to good fit indices ( 2=73.38, df =17, 2/df =4.32 (p=.04), CFI=.77, Gamma hat=.95, RMSEA=.106 (90%CI=.082-.131), SRMR=.074, AIC=111.38).

Highly suspect Items do not match expected patterns with sample of NZ university education students Not recommended Too short Possibly invalid previous analytic methods

Education and social work students being prepared for professional work in high levels of cultural, ethnic, and linguistic diversity are expected to endorse such attributes Although this study shows that a potentially important predictor of self-reported endorsement of graduate attributes is, in part, a function of personality. BUT BFI10 is not trustworthy Future research with better measures of personality is needed Would this help in recruitment and preparation of future teachers?