A common task in data science at internet companies is to develop metrics that capture aspects of the user experience. In this thesis, we are interested in systems of measurement variables without direct causal relations such that covariance is explained by unobserved latent common causes. A framework for modeling the data generating process is given by Neuro-Causal Factor Analysis (NCFA). The graphical model consists of a directed graph with edges pointing from the latent common causes to the measurement variables; its functional relations are approximated with a constrained Variational Auto-Encoder (VAE). We refine the estimation of the graphical model by developing an MCMC algorithm over Bayesian networks from which we read marginal independence relations between the measurement variables. Unlike standard independence testing, the method is guaranteed to yield an identifiable graphical model. Our algorithm is competitive with the benchmark, and it admits additional flexibility via hyperparameters that are natural to the approach. Tuning these parameters yields superior performance over the benchmark. We train the improved NCFA model on Spotify user behavior data. It is competitive with the standard VAE on data reconstruction with the benefit of causal interpretability and model identifiability. We use the learned latent space representation to characterize clusters of Spotify users. Additionally, we train an NCFA model on data from a randomized control trial and observe treatment effects in the latent space. / En vanlig uppgift för en data scientist på ett internetbolag är att utveckla metriker som reflekterar olika aspekter av användarupplevelsen. I denna uppsats är vi intresserade av system av mätvariabler utan direkta kausala relationer, så till vida att kovarians förklaras av latenta gemensamma orsaker. Ett ramverk för att modellera den datagenererande processen ges av Neuro-Causal Factor Analysis (NCFA). Den grafiska modellen består av en riktad graf med kanter som pekar från de latenta orsaksvariablerna till mätvariablerna; funktionssambanden uppskattas med en begränsad Variational Auto-Encoder (VAE). Vi förbättrar uppskattningen av den grafiska modellen genom att utveckla en MCMC algoritm över Bayesianska nätverk från vilka vi läser de obetingade beroendesambanden mellan mätvariablerna. Till skillnad från traditionella oberoendetest så garanterar denna metod en identifierbar grafisk modell. Vår algoritm är konkurrenskraftig jämfört med referensmetoderna, och den tillåter ytterligare flexibilitet via hyperparametrar som är naturliga för metoden. Optimal justering av dessa hyperparametrar resulterar i att vår metod överträffar referensmetoderna. Vi tränar den förbättrade NCFA modellen på data om användarbeteende på Spotify. Modellen är konkurrenskraftig jämfört med en standard VAE vad gäller rekonstruktion av data, och den tillåter dessutom kausal tolkning och identifierbarhet. Vi analyserar representationen av Spotify-användarna i termer av de latenta orsaksvariablerna. Specifikt så karakteriserar vi grupper av liknande användare samt observerar utfall av en randomiserad kontrollerad studie.

This measurement study examined the construct validity of the retell component of the Texas Middle School Fluency Assessment (Texas Education Agency, University of Houston, & The University of Texas System, 2008a) within a confirmatory factor analysis framework. The role of retell, provided after a one-minute oral reading fluency measure, was investigated by comparing the fit of a three-factor model of reading competence to the data collected on a diverse sample of seventh- and eighth-grade students (N=394). The final model demonstrated adequate to mediocre fit (χ2 = 97.316 {32}; CFI = 0.958; TLI = 0.941; RMSEA = .081). Results suggest that retell was a significant contributor to comprehension (Δχ2=16.652{1}, p < .001), fluency (Δχ2=10.882{1}, p = .001), and word identification (Δχ2=7.84{1}, p = .005). However, the χ2 difference was greater for comprehension, as was the factor loading for comprehension (.250, p < .001) compared to fluency (.194, p < .001) and word identification .167, p < .001). Retell did, however, have a large residual variance (.938), suggesting it did not function well as a measure of comprehension in its current state with low inter-rater reliability (K = .37). Narrative retell scores (.352, p< .001) were better predictors of comprehension than expository retell scores (from .2221 to .264, p < .001) or the combination of all three scores (Δχ2=134.261{19}; p < .001), but average retell scores produced a more parsimonious model than narrative retell scores alone (ΔAIC = 58.275; ΔBIC = 58.275). Average retell was only weakly correlated to other measures of comprehension (from r = .155 to r = .257, p < .01). However, the relationship was stronger than the relationship between retell and other measures of fluency (from r = .158 to r = .183, p < .01) or word identification (r = .132, p < .05). In addition, retell did not demonstrate differential item functioning when student characteristics (e.g., primary language, socioeconomic status, ability level) were entered as covariates, even though there were overall latent differences. / text

Background: Support for the four factor construct validity of the third edition of the Wechsler Adult Intelligence Scale (WAIS-III) has been found in clinical and non clinical populations but some studies question whether more complex models consistent with the concepts of fluid and crystallised intelligence provide a better explanation of the data. The WAIS-III is frequently used in the diagnosis of learning disability, however, previous exploratory factor analysis of data from a population with low IQ did not support the explicit four factor structure of the WAIS-III. Method: A confirmatory factor analysis of the WAIS-III was carried out on data from people with severe and significant learning disability and people with borderline IQ (IQ = 70-79). Results: The data from the borderline IQ sample and the sample with significant learning disability showed at best a weak fit to the explicit four factor models and more complex five or six factor models. However fit of the data from the sample with severe learning disability was poor for all models. Discussion: The findings show little support for the explicit four factor construct validity of the WAIS-III for people with borderline IQ or significant or severe intellectual impairment. Some support is found for the direction taken by the new Wechsler children’s and adult scales (WISC-IV & WAIS-IV) in aligning interpretation of the scales more closely to concepts such as fluid and crystallised theory. The research also suggests the cut-off point of IQ 70 is not reflective of an actual difference in cognitive profile as measured by the WAIS-III. Limitations of this study and implications for further research are also discussed.

The overall aim of the present thesis was three-fold. The first was to study occupational performance in individuals with severe mental disorders and their experiences of occupational therapy, the second to study experienced burden of family caregivers and the third to test the validity and the homogeneity of assessment tools in this area. The samples consisted of individuals with severe mental disorders participating in organised occupations (n= 112), occupational therapy records (n=64), occupational therapists working in mental health care (n=7) and family caregivers of individuals with severe mental disorders (n=256). Data were collected by questionnaires, structured and narrative interviews, observations and occupational therapy records. Individuals with severe mental disorders reported problems related to leisure and work activities and the occupational therapists recorded problems concerning how to organise and structure occupational performance. Individuals functioning on a high cognitive level experienced problems related to work and productive activities. Participation in occupational therapy strengthened their confidence in their own ability. The “Experience of Occupational Performance Questionnaire” (EOPQ) was developed from data on the experiences of women participating in occupational therapy. A principal component analysis gave seven factors with acceptable homogeneity. There is a need for assessment tools to evaluate occupational therapy. The EOPQ represents an attempt to fulfil this need. Family caregivers experienced limitations of daily activities as a burden. The ability to perform daily activities was studied from three perspectives, the individuals’, the occupational therapists’, and the experienced burden of the family caregivers. These perspectives are complementary and thus necessary for planning and implementation of individually adapted occupational therapy as well as for the evaluation of outcomes.

Previous systematic reviews indicate a lack of reporting of reliability and validity evidence in subsets of the medical education literature. Psychology and general education reviews of factor analysis also indicate gaps between current and best practices; yet, a comprehensive review of exploratory factor analysis in instrument development across the continuum of medical education had not been previously identified. Therefore, the purpose for this study was critical review of instrument development articles employing exploratory factor or principal component analysis published in medical education (2006-2010) to describe and assess the reporting of methods and validity evidence based on the Standards for Educational and Psychological Testing and factor analysis best practices. Data extraction of 64 articles measuring a variety of constructs that have been published throughout the peer-reviewed medical education literature indicate significant errors in the translation of exploratory factor analysis best practices to current practice. Further, techniques for establishing validity evidence tend to derive from a limited scope of methods including reliability statistics to support internal structure and support for test content. Instruments reviewed for this study lacked supporting evidence based on relationships with other variables and response process, and evidence based on consequences of testing was not evident. Findings suggest a need for further professional development within the medical education researcher community related to 1) appropriate factor analysis methodology and reporting and 2) the importance of pursuing multiple sources of reliability and validity evidence to construct a well-supported argument for the inferences made from the instrument. Medical education researchers and educators should be cautious in adopting instruments from the literature and carefully review available evidence. Finally, editors and reviewers are encouraged to recognize this gap in best practices and subsequently to promote instrument development research that is more consistent through the peer-review process.

More than a decade of research has indicated that individuals who have experienced traumatic events may report positive psychological changes as a result of their struggle to cope with the impact of the event. This cluster of changes has been labelled posttraumatic growth (PTG). Several measurement instruments have been developed to quantify such growth, one of the most widely used being the Posttraumatic Growth Inventory (PTGI) developed by Tedeschi and Calhoun (1996). The objective of the present study was to conduct a confirmatory factor analysis (CFA) of the PTGI on data yielded by a South African sample, to determine if the hypothesised factor structure showed appropriate fit in this context. Analyses were performed with three hypothesised models: the five factor model, the three factor model and the model of posttraumatic growth as a unitary factor. Whilst the results of the CFA did not provide overt support for the adequate fit of the models tested, it did yield a profile of PTG in the South African sample that is similar to that of samples from other countries of origin. The findings implied that the factorial validity of the PTGI is unclear and the inventory may require modification for use in the South African context. This highlighted the need for further investigation in order to provide a comprehensive exploration of the PTGI and its applicability in this context. This is in line with the current focus on the socio-cultural and contextual elements that may influence the perception of PTG.

Ensuring that teachers have high mental health literacy (MHL) is paramount, particularly because the prevalence rate for mental illness among children and adolescents is 10-20% (Mental Health Commission of Canada, 2013). Currently, there are few programs designed to increase the MHL of teachers, and the research base examining these is scant. As a result, more research is required to find effective methods of improving MHL among teachers. The current study explored the influence of a new, interactive MHL learning module on the MHL levels of B.Ed. candidates. A questionnaire largely adapted from existing measures was developed and factor analysis was conducted. Pre- and post-module MHL scores were analyzed using descriptive approaches and a repeated-measures MANOVA. Results indicate that pre-service teachers have high knowledge and non-stigmatizing beliefs in relation to MHL which did not change post- module completion; self-efficacy levels did show a small, non-significant increase.

The 24 items of the original English version of the Social Justice Scale (Torres-Harding et al., 2012) were translated into Korean by four translators, who discussed and agreed upon consensus versions. Four different translators then back translated this version into English. The resulting Korean version of SJS (K-SJS) was completed by 537 adult native Korean speakers. Confirmatory factor analysis, exploratory factor analysis, and multi-group confirmatory factor analysis indicated that the K-SJS has high internal consistency, factors appropriately, fits the original model well, and demonstrates invariance across Korean men and women. Structural equation modeling indicated that the effects of attitude, perceived behavioral control, and subjective norms on behavioral intentions were positive and significant. In short, the K-SJS showed acceptable reliability and validity based on a large sample of South Korean adults and shows promise as a new tool to study social justice attitudes among Korean speakers.

Ringo M.H. Ho. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1997. / Includes bibliographical references (leaves 75-81). / ACKNOWLEDGMENT --- p.2 / ABSTRACT --- p.3 / TABLE OF CONTENTS --- p.5 / LIST OF TABLES --- p.8 / Chapter CHAPTER 1 --- PREVIOUS STUDIES ON USING PROCRUSTES ROTATION TO ASSESS FACTORIAL INVARIANCE --- p.10 / Factorial invariance problem --- p.10 / Procrustes rotation with congruent coefficient as a way to test factorial invariance --- p.11 / Quantifying the Procrustes fit --- p.14 / Outline of the present study --- p.15 / Chapter CHAPTER 2 --- A CRITICAL EVALUATION OF THE PERMUTATION METHOD --- p.18 / Introduction --- p.18 / Method --- p.19 / Results and Discussions --- p.21 / Chapter CHAPTER 3 --- BOOTSTRAP TESTING PROCEDURE FOR A FULLY SPECIFIED TARGET --- p.24 / Introduction --- p.24 / A brief introduction to the bootstrap procedure --- p.24 / The bootstrap testing procedure for a fully specified target --- p.26 / Method --- p.28 / Results and Discussions --- p.28 / Chapter CHAPTER 4 --- BOOTSTRAP TESTING FOR A PARTIALLY SPECIFIED TARGET --- p.33 / Introduction --- p.33 / The bootstrap testing procedure for a partially specified target --- p.36 / Method --- p.38 / Quantifying the fit - congruence coefficients for the partial target rotation --- p.39 / Results and Discussions --- p.40 / Chapter CHAPTER 5 --- FURTHER EXTENSIONS OF THE BOOTSTRAP METHOD --- p.45 / Introduction --- p.45 / First extension - when correlation matrix is used --- p.45 / The modified bootstrap procedure --- p.45 / Method --- p.48 / Results and Discussions --- p.48 / Second extension - when raw data of the target sample is not available --- p.49 / The conditional bootstrap procedure for a fully specified target --- p.49 / Method --- p.50 / Results and Discussions --- p.51 / Chapter CHAPTER 6 --- THREE REAL EXAMPLES --- p.54 / Example 1 - Testing factorial invariance of CPAI between two random split samples --- p.54 / Results --- p.55 / Example 2 - Testing factorial invariance of CPAI between Chinese males and females --- p.56 / Results --- p.57 / Example 3 - Cross-cultural comparison of the Big Five Model between U. S. and Chinese samples --- p.58 / Results --- p.59 / Chapter CHAPTER 7 --- CONCLUSIONS --- p.62 / Practical remarks on the bootstrap procedure --- p.62 / A note on the transformation on the sample for constructing correct resampling space --- p.64 / Remarks on utilizing the congruence coefficients --- p.65 / How good are the congruence coefficients in detecting discrepancy between two factor structures? --- p.68 / Rule of thumb for factor congruence coefficient in checking factor replicability --- p.68 / Sample size requirement --- p.69 / Limitations of the present study --- p.70 / Direction of future studies --- p.71 / Concluding remarks --- p.73 / REFERENCES --- p.75 / NOTES --- p.82 / APPENDIX1 --- p.83 / TABLES 1 TO TABLES17 --- p.84

by Yuen Wai-ying. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1998. / Includes bibliographical references (leaves 50-51). / Abstract also in Chinese. / Chapter Chapter 1 --- Introduction --- p.1 / Chapter Chapter 2 --- Estimation of Factor Scores in a Three-level Factor Analysis Model / Chapter 2.1 --- The Three-level Factor Analysis Model --- p.5 / Chapter 2.2 --- Estimation of Factor Scores in Between-group --- p.7 / Chapter 2.2.1 --- REG Method --- p.9 / Chapter 2.2.2 --- GLS Method --- p.11 / Chapter 2.3 --- Estimation of Factor Scores in Second Level Within-group --- p.13 / Chapter 2.3.1 --- REG Method --- p.15 / Chapter 2.3.2 --- GLS Method --- p.17 / Chapter 2.4 --- Estimation of Factor Scores in First Level Within-group / Chapter 2.4.1 --- First Approach --- p.19 / Chapter 2.4.2 --- Second Approach --- p.24 / Chapter 2.4.3 --- Comparison of the Two Approaches in Estimating Factor Scores in First Level Within-group --- p.31 / Chapter 2.5 --- Summary on the REG and GLS Methods --- p.35 / Chapter Chapter 3 --- Simulation Studies / Example1 --- p.37 / Example2 --- p.42 / Chapter Chapter 4 --- Conclusion and Discussion --- p.48 / References --- p.50 / Figures --- p.52