The research design is quasi-experimental and studies a naturally occurring intervention on a self-selected sample. The intervention is visual arts training (minimum of 9 hours/week). The outcome measure is spatial reasoning in the arts and in geometry.The treated vs. untreated control conditions test for whether any kind of intensive treatment has the same effects. See description of control conditions under Setting.
The study is designed to test the causal hypothesis that instruction in visual arts strengthens spatial reasoning which can then be applied to geometric reasoning. The research design will also includes the identification of associative/correlational relationships within groups. The study also has a qualitative component in that we will be conducting in-depth “loud thinking” interviews with a small sample of students in each testing condition.
The study uses a battery of measures. A geometry visualization measure consists of release items from TIMSS, PISA, NAEP and original items created in conjunction with advisory board members. An art visualization measure, created in conjunction with artist/art educator consultants, assesses aspects of artistic visual/spatial reasoning: identifying the light source, drawing shadows, identifying reflections, imagining new perspectives, completing a scene, identifying the rhythm and shapes underlying compositions, and stretching/compressing compositions. Three standardized spatial reasoning tests provide a validity check for the spatial nature of the art and geometry items: a water level test, a paper folding test, and a mental rotation test. Two control tests are also included: the vocabulary subtest of Kaufman Brief Intelligence test, and the Reading the Mind in the Eyes test. The project also uses a custom-designed self-completion questionnaire for all participants
We will analyze a set of hypotheses about the relative performance of treatment, treated control, and untreated control groups using factor analyses, correlational analyses, analyses of variance, and SEM analyses. We will develop scoring rubrics and coding categories for analyzing the qualitative data that we will obtain from the loud-thinking interviews.