Disciplinary Attentiveness to Student Ideas in Physics (DASI-P) Dataset

This dataset represents a de-identified analytic dataset from a pilot administration of the Disciplinary Attentiveness to Student Ideas in Physics (DASI-P) instrument, designed for research and program-level evaluation rather than individual assessment. Data collection. Data were collected via an online survey administered in Qualtrics to physics educators. Respondents were randomly assigned up to 10 tasks from a pool of 23 tasks to manage survey length. Each task included five prompts: identification of task intent (T), interpretation of Student A’s response (SA), proposed instructional next step for Student A (TA), interpretation of Student B’s response (SB), and proposed instructional next step for Student B (TB). Randomized task assignment resulted in structural missingness that reflects survey design rather than item nonresponse. Raw data and processing. Raw survey responses were initially collected as free-response text. These responses were coded by the research team into three categories—Strong, Weak, and Vague/Generic—based on alignment with disciplinary intent, interpretation of student thinking, and appropriateness of instructional response. The public dataset includes only the coded analytic data; raw free-response text, handwritten student artifacts, and Qualtrics administrative metadata (e.g., IP addresses, timestamps) are excluded to protect participant privacy. Dataset structure and attributes. Each column corresponds to a coded response for a specific question type within a specific physics task. Column names follow a structured naming convention combining a task code and a question-type code (e.g., FSP_SB = interpretation of Student B’s response for the Forces Skateboard Push task). Responses are categorical (Strong / Weak / Vague) and are not assumed to be strictly ordinal; category functioning is treated as an empirical question during analysis. Analysis context. The dataset supports analyses of instrument functioning, item performance, and response-category behavior, including exploratory psychometric and item-level triage analyses. It is intended to inform instrument refinement and short-form identification. No formulas are used to generate derived variables in the public dataset.