Beyond Computations-Research Data
Description
Research Hypothesis This study hypothesized that Indonesian senior high school students' difficulties in solving Systems of Linear Equations in Three Variables (SPLTV) story problems are primarily located at the comprehension and transformation stages of Newman's Error Analysis (NEA), and that error patterns would qualitatively differ across student ability levels. Data Collection Data were collected from 81 tenth-grade students at a public senior high school in West Sumatra, Indonesia. Students completed a written test of two SPLTV story problems validated by five experts (two mathematics lecturers, two senior mathematics teachers, and one Bahasa Indonesia language lecturer). Six students (two high-ability, two medium-ability, two low-ability) were selected for semi-structured interviews. All interviews were audio-recorded and transcribed verbatim. What the Data Shows The dataset includes: Student Test Scores: Numerical scores for 81 students Students' Written Answers: Anonymized responses to two test problems Interview Transcripts: Verbatim transcripts of six semi-structured interviews Notable Findings Analysis of 81 test scripts revealed 127 errors distributed across NEA stages: Comprehension (22), Transformation (26), Process Skills (41), Encoding (8), and No Answers (30). No Reading Errors were identified. Stratified analysis revealed distinct error patterns: High-ability students: Fragile conceptual understanding and carelessness under cognitive load Medium-ability students: Linguistic complexity barriers in Bahasa Indonesia word problems Low-ability students: Compounded cognitive deficits and affective barriers (low self-efficacy, learned helplessness) Data Interpretation and Use This dataset can be used to replicate NEA coding protocols, analyze error distributions, examine the relationship between linguistic complexity and mathematical comprehension, and develop targeted instructional interventions. The qualitative interview data provide rich insights into student reasoning. Ethical Considerations All data have been anonymized. Informed consent was obtained from all participants and their parents/guardians.
Files
Steps to reproduce
Data and Methods Outline: Beyond Computation Research Design Qualitative descriptive approach using Newman's Error Analysis (NEA) framework to diagnose errors in SPLTV story problems. Participants 81 tenth-grade students (aged 15-16) from a public senior high school in West Sumatra, Indonesia 6 selected for interviews (2 high, 2 medium, 2 low ability) using purposive sampling Instruments Written Test: Two SPLTV story problems validated by 5 experts: 2 mathematics lecturers, 2 senior mathematics teachers, 1 Bahasa Indonesia lecturer Assessed: content validity, construct validity, linguistic clarity Semi-Structured Interviews: Audio-recorded, transcribed verbatim Data Collection Procedures Institutional permissions and informed consent obtained Written test administered (60 minutes) Test scoring and candidate identification Interviews conducted (15-30 minutes each) Interviews transcribed verbatim Data Analysis Workflow Stage 1: Test Analysis (Error Classification) Three researchers independently coded 20 test scripts using NEA framework Inter-rater reliability: 89% agreement 127 errors categorized into five NEA stages (Reading, Comprehension, Transformation, Process Skills, Encoding) Stage 2: Interview Analysis (Thematic Analysis) Data reduction: Deductive coding using NEA stages + inductive coding for emergent themes Data display: Thematic matrices organized by student and ability level Conclusion drawing: Pattern identification and verification Trustworthiness Strategies Credibility: Data triangulation, expert validation, member checks Dependability: Audit trail of all decisions Confirmability: Peer debriefing, reflexivity Software Used Mendeley (reference management) SciSpace (literature discovery) Microsoft Word/Excel (documentation) Reproduction Instructions Use provided test problems (see article Appendix) Validate test with expert panel (minimum 3 experts) Administer to Grade 10 students (60 minutes) Apply NEA coding framework with established reliability Conduct semi-structured interviews with purposive sample Perform thematic analysis using deductive and inductive coding Contact Adri Nofrianto: adrinofrianto@gmail.com