
Education
Enhancing students’ attitudes towards statistics through innovative technology-enhanced, collaborative, and data-driven project-based learning
A. Cujba and M. Pifarré
This exciting research by Andreea Cujba and Manoli Pifarré explores how innovative teaching methodologies, particularly technology-enhanced, collaborative learning, positively influence students' attitudes towards statistics, revealing a promising shift towards decreased anxiety and increased enthusiasm for stats. Discover the transformative power of teaching in education!
~3 min • Beginner • English
Introduction
The paper addresses the growing need for data analytics skills in a data-rich, AI-driven world and the challenge that many secondary students hold negative attitudes and anxiety toward statistics, which hinders learning. Prior research links positive attitudes toward statistics with better academic outcomes, yet many learners fail to see statistics’ real-life relevance. The study proposes that combining project-based learning (PBL), data analytics (DA), collaborative work, and technology can create a meaningful learning environment that improves attitudes. The research question asks about the effects of a collaborative, technology-enhanced, data-driven project-based intervention on students’ attitudes toward statistics. The authors hypothesize that, relative to a traditional curriculum, the intervention will: improve global attitudes (H1), reduce anxiety (H2), increase affect (H3), and improve attitudes toward learning statistics with technology (H4).
Literature Review
Attitudes are multifaceted constructs comprising beliefs, feelings, and behavioral predispositions, shaped through experience and influencing future behavior. In statistics education, attitudes are emotional reactions to learning experiences and expectations about the subject; math anxiety is a common negative affective response. Research suggests innovative methods can promote positive attitudes, notably student-centered learning, PBL with real-world, open-ended problems, DA skills, collaborative learning, and interactive technologies. PBL contextualizes statistics in authentic challenges, heightens motivation, and can improve attitudes and affect. DA focuses on understanding and interpreting data to solve real problems rather than rote procedures, reinforcing active, inquiry-based learning. Technology (e.g., TinkerPlots, CODAP, Fathom) enables interactive visualization, exploration of large datasets, modelling, pattern identification, and graph creation, which supports reasoning and motivation. Studies report technology’s role in fostering collaboration, engagement, and statistical literacy. Collaborative learning, scaffolded by explicit ground rules and exploratory talk, enhances communication, group functioning, and positive attitudes. The literature indicates that integrating PBL, DA, technology, and collaboration may create synergistic benefits for attitudes toward statistics, motivating the present study.
Methodology
Design: A quasi-experimental study within the EU ERASMUS+ SPIDAS project compared an experimental group (EG) receiving a collaborative, technology-enhanced, data-driven project-based intervention with a control group (CG) receiving traditional instruction. Participants: 174 grade 8 students (13–14 years) from two Spanish publicly funded private schools: EG n=110 (52.7% girls), CG n=64 (53.12% girls), with similar medium socioeconomic profiles and comparable academic achievement (per National Test of Basic Skills). Prior statistical knowledge was limited. Instrument: A 16-item, 4-point Likert (1=strongly disagree to 4=strongly agree) questionnaire assessing attitudes toward statistics with technology, validated in Spanish via exploratory factor analysis (varimax) revealing three factors: Anxiety, Learning statistics with technology, and Affect. Internal consistency: overall α=0.83 (Anxiety α=0.83; Technology α=0.76; Affect α=0.77). Negatively worded items (anxiety) were reverse-scored. The questionnaire was administered pre- and post-intervention. Experimental intervention (SPIDAS): 30 hours over 2 months. Students worked in small groups (3–4) on a real-life project, “How does the weather affect our lives?”, combining classroom and outside work (synchronous collaboration via Google Drive). Three pedagogical axes: (1) Data-driven PBL with an explicit DA cycle inspired by PPDAC and informal statistical inference: Define the problem (group-chosen research question/hypothesis), Consider data (design/validate surveys; collect and check data quality), Explore data (use CODAP for visualization, pattern-finding, relationships, certainty statements), Draw conclusions (answer RQ/test hypotheses), Make decisions (propose actions; communicate via infographics/video), Evaluate courses of action (reflect on methods, results, and next steps; write a report). (2) Collaborative learning: Strategies from the Thinking Together program—role reflection, norms/behaviors that promote collaboration, and development of effective ground rules to support exploratory talk and co-construction. (3) Technology: CODAP supported interactive data exploration, graphing, and interpretation; Google Docs/Slides enabled synchronous, multi-user workspaces for discussion, co-construction, planning, and documentation. Teacher role blended coaching and guiding data-analytic processes while students led inquiry. Control intervention: 2 months of teacher-centered lectures covering the same statistical literacy topics (mean, median, mode, range, variability, variable types, frequency, proportional reasoning, graph reading, sample, population). Students worked individually on routine exercises, often at home, using teacher-provided (non-contextualized) datasets. Excel was used primarily for calculations and graphing, requiring procedural knowledge and emphasizing computations over interpretation. Data analysis: Normality was tested with Shapiro–Wilk; due to non-normality, non-parametric tests were applied. Intragroup pre–post differences were evaluated with Wilcoxon signed-rank tests. Intergroup differences (EG vs CG) at pre and post were assessed with Mann–Whitney U tests. Effect sizes (reported as r in tables) accompanied significance tests.
Key Findings
Intragroup (Wilcoxon) results: • Experimental group (EG) showed significant improvements: Global Attitude toward statistics p<0.001, r=0.42; Anxiety factor p<0.001, r=0.40 (anxiety decreased; reverse-scored items increased); Technology factor p<0.001, r=0.37. Affect showed a positive trend but was not significant p=0.299, r=0.10. • Control group (CG) showed no significant changes in any variable (Global Attitude p=0.748; Anxiety p=0.777; Affect p=0.571; Technology p=0.122). Intergroup (Mann–Whitney U) results: • Pre-test differences: EG > CG in Global Attitude (p=0.010, r=−0.20) and Affect (p=0.001, r=−0.25); Anxiety (p=0.064) and Technology (p=0.379) not significantly different. • Post-test differences favored EG with statistical significance on all variables: Global Attitude Z=−4.397, p<0.001, r=−0.33; Anxiety Z=−4.530, p<0.001, r=−0.34; Affect Z=−4.569, p<0.001, r=−0.35; Technology Z=−1.974, p=0.048, r=−0.15. Descriptive pre/post means (Fig. 6): • Global Attitude (sum of 16 items): EG 48.00→51.00; CG 44.73→45.61. • Anxiety factor: EG 15.72→17.11; CG 14.64→14.97. • Affect factor: EG 13.95→14.25; CG 12.25→11.83. • Technology factor: EG 18.33→19.68; CG 17.84→18.81. Interpretation: The SPIDAS intervention significantly improved students’ overall attitudes, reduced anxiety, and increased attitudes toward learning with technology; affect showed a within-group positive trend and significant intergroup post differences. The CG’s traditional instruction produced no significant improvements. Overall, findings support H1, H2, and H4, and provide intergroup evidence consistent with H3.
Discussion
The findings indicate that integrating data-driven PBL, collaborative learning, and interactive technology (CODAP and Google collaborative tools) creates a meaningful, student-centered environment that improves attitudes toward statistics. By situating statistics within real-life, student-selected questions and emphasizing interpretation and decision-making, the EG developed more positive global attitudes, experienced reduced anxiety, and reported improved views of learning with technology. The structured collaborative strategies likely enhanced group functioning, dialogue, and co-construction, further supporting positive attitudes. In contrast, the traditional, computation-focused CG with less contextualization and individual work offered limited opportunities for relevance, agency, and dialogue, aligning with literature that such conditions dampen affect and engagement. Although within-group affect gains for EG were not statistically significant, the intergroup post differences and CG’s decline suggest that SPIDAS buffered against negative affect and trended toward improvement. The modest affect gains may reflect the initial cognitive and organizational demands of first-time PBL and DA experiences; repeated or longer exposures could yield stronger affective benefits. Technology’s role was salient: CODAP’s visualization and ease of exploration appear to shift effort from computation to reasoning, supporting positive attitudes and skill development. Overall, the combined pedagogical design addresses known barriers (anxiety, perceived irrelevance) and advances a practical model for fostering attitudes essential to statistical learning.
Conclusion
The study demonstrates that a collaborative, technology-enhanced, data-driven project-based approach can improve secondary students’ attitudes toward statistics, reduce anxiety, and enhance attitudes toward learning with technology, while trending positively for affect relative to traditional instruction. By engaging students in authentic data inquiry—formulating questions, collecting and analyzing real data, drawing inferences, and communicating decisions—the SPIDAS design cultivates meaningful engagement and statistical thinking. Contributions include evidence from a real-classroom, multi-week intervention that unifies PBL, DA, collaborative strategies, and interactive tools. Future research should extend the duration and frequency of such interventions to strengthen affective gains, examine how low (productive) levels of anxiety function during statistical learning, consider socioeconomic status as a moderating variable, and further validate and generalize the attitude questionnaire across broader student populations and contexts.
Limitations
• Novelty and cognitive load: This was students’ first intensive experience with PBL and CODAP; the high cognitive demands may have tempered affect gains. Longer, longitudinal interventions may mitigate this. • Residual anxiety: Despite reductions, some students still reported anxiety; future work could use qualitative methods to capture potentially productive (motivational) levels of anxiety. • Socioeconomic status: Although schools were similar, SES should be analyzed explicitly as an independent variable in future research. • Measurement generalizability: The attitudes questionnaire, while showing good internal consistency and exploratory validity in Spanish, requires further validation, larger samples, and confirmatory analyses to establish external validity across contexts and educational levels.
Related Publications
Explore these studies to deepen your understanding of the subject.