2019 Session III: Rosa M. Ailabouni

Rebecca Berenbon, Christa Winkler
berenbon.1@osu.eduwinkler.99@osu.edu
Educational Studies, Quantitative Research, Evaluation, and Measurement
Jerome D’Agostino, Advisor

So Much Data, So Little Time: Validation of a Scale that Measures Researchers’ Data Reuse Behaviors

This study investigated the validity of a survey measuring scientists’ attitudes toward data reuse. Several techniques, including parallel analysis, exploratory factor analysis, and Rasch analysis, were used to study the psychometric properties of the survey. Implications for future modifications and use of the data reuse measure are discussed.


Meng-Ting Lo, James Uanhoro
lo.194@osu.eduuanhoro.1@osu.edu
Educational Studies, Quantitative Research, Evaluation, and Measurement
Ann O’Connell, Advisor

When Does the Linear Regression Model Work Well for an Ordinal Outcome?

The current study focuses on comparing performance of linear regression and ordinal regression models by examining their statistical power and mean square error (MSE) of predicted values underlying different study conditions of sample sizes, distribution and the number of categories for an ordinal outcome, and the probability of a binary predictor.


Susan Mauck
mauck.2@osu.edu
Educational Studies, Quantitative Research, Evaluation, and Measurement
Ann O’Connell, Advisor

Measuring Fundamental Applied Statistics Knowledge–Development of a Formative Assessment for Graduate Students and Quantitative Methods Instructors

A test for graduate students to measure their fundamental applied statistics knowledge has been developed and had its use validated using a rigorous test development study design. The process to develop the test and the statistical analyses of the large-sample pilot test data (n = 412) will be described.


Meingold Chan
chan.742@osu.edu
Human Sciences, Human Development and Family Service (HDFS)
Xin Feng, Advisor

Application of Machine Learning: Analyzing Family Emotional Climate with Sentiment Analysis using R Packages

The current study shows how to use sentiment analysis (SA), a popular application of machine learning, to extract emotional component to assess family emotional climate. This study compares the performance (accuracy rate of predicting manual classification of sentiment) of two R packages for SA using transcripts of family daily conversation.