Student attitude toward STEM: Development of an instrument for high school STEM-based programs

Expansive amounts of money and time have been provided in the hopes that STEM based programs will boost student interest and abilities related to STEM. Financiers include national government and industrial organizations. However, these investments have yielded little results as demonstrated by the continued reports being disseminated each year demanding greater STEM investment and results. The development of an instrument that can accurately measure student attitude toward STEM is crucial to STEM-based programs, their intended outcomes, and the companies that aid in their function as well as hope to reap the benefits from their products.;The intent of this study was to develop an instrument to measure the current level of attitude that students' exhibit toward STEM and STEM education. A variation of the Concerns-Based Adoption Model, Taxonomy of Education Objectives---Handbook II, and other pertinent instruments were utilized as sources of inspiration for the development of the instrument. The developed and selected items were submitted to a panel of experts representative of STEM and STEM education. Initial pilot testing refined the instrument through principal components analysis and Cronbach's alpha coefficients. Three principal components were identified by the researcher---interest, ability, and value. The identified principal components aligned well with reviewed attitudinal instruments. Reliability coefficients were strong for each of the identified principal components---coefficients were above .70 alpha.;Concurrent validity was also established for the pilot study by comparing the data analysis results obtained from the student attitude toward STEM instrument with the results from a semantic differential attitudinal instrument. The example of concurrent validity provided by the Pearson product moment correlation between the two attitudinal instruments only applied to the student attitude toward STEM instrument used for the pilot study. Both attitudinal instruments were included in the pilot study instrument packet provided to the high school students. A statistically significant moderately positive correlation was demonstrated between the two instruments as indicated by the Pearson product moment correlation.;A student focus group was used to review the pilot instrument and the instrument items. The students provided valuable face and content validity regarding high school student interpretation of the instrument. Results of the combined analyses led to revisions of the instrument prior to a larger comparative study---a known-group comparison. A self-identified STEM-based high school program and a conventional college-preparatory program were compared in the known-group comparison. Principal components analysis and Cronbach's alpha procedures were again applied to the data collected from the revised instrument. The two samples were compared using three distinct independent variables---educational location (type of school), grade level (ninth-grade and eleventh-grade), and gender (male and female). Each independent variable was analyzed for each of the researcher identified principal components: interest, ability, and value.;Hypotheses were designed and tested for each of the independent variables. The hypotheses were used in an attempted to provide another example of construct validity for the revised student attitude toward STEM instrument. MANOVA procedures along with follow up univariate analyses and non-parametric analyses procedures were applied to the student responses to the revised STEM attitudinal instrument. Assumptions regarding homogeneity of variance and normality were also carried out for the statistical procedures. Homogeneity of variance was maintained for the majority of independent variable reviewed with violations being specific to certain content areas or singular attitudinal levels. Normal distributions were not demonstrated by the majority of the independent variable samples; the distributions were primarily negatively-skewed.;According to the results of the data analyses, the male students indicated a statistically significant more positive attitude toward STEM when compared to the female students for the independent variable of gender. The statistical significance was demonstrated specifically for the content areas of technology and engineering. The results of the data analysis supported the proposed hypothesis for the content area of technology and engineering and provided the revised student attitude toward STEM instrument with an example of construct validity. The data analyses of the independent variables of school and grade level did not provide statistically significant results. The null hypotheses of for these two independent variables were retained.;Based upon extensive review of the varied data analysis procedures implemented, the student attitude toward STEM instrument demonstrated positive examples of validity and reliability. The results of the exploratory research study are promising for the development of a useful attitudinal instrument as well as powerful tool for STEM education. Future research with larger and more varied sample sizes will aid in the development and refinement of the student attitude toward STEM instrument. This study and the resulting attitudinal instrument are imperative steps toward increasing the accountability and design of quality STEM educational programs.