Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 126 of the 127 references it cites were found in a registry or an index; 4 of the 45 sources its claims cite were checked for support.

Synthesizing Results From Empirical Research on Computer-Based Scaffolding in STEM Education

Open access
Belland, Brian R., Walker, Andrew E., Kim, Nam Ju, Lefler, Mason · 2017 · Review of Educational Research

Abstract

Computer-based scaffolding assists students as they generate solutions to complex problems, goals, or tasks, helping increase and integrate their higher order skills in the process. However, despite decades of research on scaffolding in STEM (science, technology, engineering, and mathematics) education, no existing comprehensive meta-analysis has synthesized the results of these studies. This review addresses that need by synthesizing the results of 144 experimental studies (333 outcomes) on the effects of computer-based scaffolding designed to assist the full range of STEM learners (primary through adult education) as they navigated ill-structured, problem-centered curricula. Results of our random effect meta-analysis (a) indicate that computer-based scaffolding showed a consistently positive (ḡ = 0.46) effect on cognitive outcomes across various contexts of use, scaffolding characteristics, and levels of assessment and (b) shed light on many scaffolding debates, including the roles of customization (i.e., fading and adding) and context-specific support. Specifically, scaffolding's influence on cognitive outcomes did not vary on the basis of context-specificity, presence or absence of scaffolding change, and logic by which scaffolding change is implemented. Scaffolding's influence was greatest when measured at the principles level and among adult learners. Still scaffolding's effect was substantial and significantly greater than zero across all age groups and assessment levels. These results suggest that scaffolding is a highly effective intervention across levels of different characteristics and can largely be designed in many different ways while still being highly effective.

Overview

CoChat Summary

TL;DR

This comprehensive random-effects meta-analysis of 144 experimental studies (333 outcomes) found that computer-based scaffolding in STEM education produced a consistently positive effect on cognitive outcomes (ḡ = 0.46). Effects were substantial across learner populations, contexts, and assessment levels, with no significant moderation by context-specificity, scaffolding change, or scaffolding logic.

Key findings

  • Computer-based scaffolding significantly improved cognitive outcomes versus no scaffolding: overall ḡ = 0.46, z = 18.19, p < .01, across 333 outcomes from 144 studies.
  • Assessment-level effects were concept ḡ = 0.40, principles ḡ = 0.51, and application ḡ = 0.44; scaffolding’s influence was greater at the principles level than at the concept level (z = 2.17, p < .05).
  • No statistically significant differences were found for context-specific versus generic scaffolding (z = −0.281, p = .778), scaffolding change/fading (p > .05), or scaffolding logic (p > .05).
  • Effects were significantly greater than zero across education levels, but adult learners had the highest estimate based on only one outcome; traditional versus underperforming learners differed significantly (z = 2.29, p < .05). Motivation scaffolding (n = 3), design-based learning (n = 3), and adult population (n = 1) estimates were not significantly greater than zero, with caution due to small samples.

Methods

  • Meta-analysis of studies published January 1, 1993–December 31, 2014, with a control condition receiving an educational intervention but no scaffolding, cognitive outcomes, sufficient effect-size information, Wood et al. (1976) scaffolding, and ill-structured STEM problems. Searches covered databases including Education Source, PsychINFO, ERIC, PubMed, IEEE, and Google Scholar, plus hand searches and reference-list searches.
  • Four coders independently coded studies, reached consensus, and achieved Krippendorff’s alphas greater than .667; 333 data points were used for interrater reliability.
  • Random-effects meta-analysis used Hedges’ g and STATA 14 metan; publication bias was assessed via funnel plot, Egger’s regression, and trim-and-fill. Five outlier outcomes and three associated studies were removed, reducing 338 outcomes/147 studies to 333 outcomes/144 studies. Effect-size dependency was addressed with composites (515 to 333 outcomes) and robust variance estimation.

Limitations

  • The paper states meta-analyses cannot include all empirical research; qualitative studies and quantitative studies without control groups were excluded, so effect sizes do not reflect all empirical research on computer-based scaffolding.
  • The coding scheme could not perfectly reflect every construct; it assigned a single value for scaffolding logic, and codes such as “none,” “traditional,” or “problem solving” may indicate lack of description as much as positive identification.
  • Several moderator estimates relied on very small samples—motivation scaffolding (n = 3), design-based learning (n = 3), and adult population (n = 1)—producing wide confidence intervals and warranting caution; no primary research on preschool populations met inclusion criteria.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3102/0034654316670999

Cites

126 of 126 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1. Search and exclusion process. k refers to number of studies, and n refers to number of outcomes. The final set of 144 included studies were associated with 333 outcomes.
Figure 1. Search and exclusion process. k refers to number of studies, and n refers to number of outcomes. The final set of 144 included studies were associated with 333 outcomes.CC BY-NC · PubMed Central
Figure 2. Funnel plot with pseudo 95% confidence limits.
Figure 2. Funnel plot with pseudo 95% confidence limits.CC BY-NC · PubMed Central
Figure 3. Comparison of effect size according to education level. n refers to the number of outcomes.
Figure 3. Comparison of effect size according to education level. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 4. Comparison of effect size according to education population. n refers to the number of outcomes.
Figure 4. Comparison of effect size according to education population. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 5. Comparison of effect size according to context of scaffolding use. n refers to the number of outcomes. The effect of design-based learning is not statistically greater than zero, p = .14. Project-based learning (ḡ = 1.33) has an estimate and confidence interval so high that it does not show on our −0.8 to 0.8 scale.
Figure 5. Comparison of effect size according to context of scaffolding use. n refers to the number of outcomes. The effect of design-based learning is not statistically greater than zero, p = .14. Project-based learning (ḡ = 1.33) has an estimate and confidence interval so high that it does not show on our −0.8 to 0.8 scale.CC BY-NC · PubMed Central
Figure 6. Comparison of effect size according to assessment level. n refers to the number of outcomes.
Figure 6. Comparison of effect size according to assessment level. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 7. Comparison of effect size according to scaffolding change. n refers to the number of outcomes.
Figure 7. Comparison of effect size according to scaffolding change. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 8. Comparison of effect size according to the basis by which scaffolding was added, faded, or added/faded. n refers to the number of outcomes.
Figure 8. Comparison of effect size according to the basis by which scaffolding was added, faded, or added/faded. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 9. Comparison of effect size according to scaffolding scope. n refers to the number of outcomes.
Figure 9. Comparison of effect size according to scaffolding scope. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 10. Comparison of effect size according to scaffolding intervention. n refers to the number of outcomes.
Figure 10. Comparison of effect size according to scaffolding intervention. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 11. Comparison of effect size according to study design. n refers to the number of outcomes.
Figure 11. Comparison of effect size according to study design. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 12. Comparison of effect size according to reliability reporting. n refers to the number of outcomes.
Figure 12. Comparison of effect size according to reliability reporting. n refers to the number of outcomes.CC BY-NC · PubMed Central
Figure 13. Comparison of effect size according to validity reporting. n refers to the number of outcomes.
Figure 13. Comparison of effect size according to validity reporting. n refers to the number of outcomes.CC BY-NC · PubMed Central

Verification

Claims the paper makes

4 Valid · 2 Not valid · 30 Not checked · 74 Own results and interpretations

59 more claims

Create a free account to see all 110 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 18
  • It may state two separate claims as one. 32
  • The quote only partly states this claim. 2

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Consistentz = 18.19reported p < .01 · recomputed p < .001

    “z = 18.19, p < .01”

  • Consistentz = 2.29reported p < .05 · recomputed p = .022

    “z = 2.29, p < .05”

  • Consistentz = 4.69reported p < .01 · recomputed p < .001

    “z = 4.69, p < .01”

  • Consistentz = 5.09reported p < .01 · recomputed p < .001

    “z = 5.09, p < .01”

  • Consistentz = 5.36reported p < .01 · recomputed p < .001

    “z = 5.36, p < .01”

5 of 13 · the rest aren’t loaded

8 more recomputed tests

Create a free account to see every test the Stats check recomputed.

See all tests — free account

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

3 of 3

Citations

Cites

Show all 127 references

Research works printed but not linked: 13

  1. Alias N. A., 2012, Journal of Educational Technology & Society, 15, 137
  2. Belland B. R., 2015, Educational Technology and Society, 18, 183
  3. Brush T., 2001, Journal of Educational Multimedia and Hypermedia, 10, 333
  4. Collins A., 1989, Knowing, learning and instruction: Essays in honor of Robert Glaser, 453
  5. Conati C., 2000, International Journal of Artificial Intelligence in Education, 11, 389
  6. Hannafin M., 1999, Instructional design theories and models: Volume II. A new paradigm of instructional theory, 115
  7. Knowles M. S., 1984, Andragogy in action. Applying modern principles of adult education
  8. 10.1111/jca.2009.25.issue-410.1111/j.1365-2729.2009.00313.x · doi:10.1111/jca.2009.25.issue-410.1111/j.1365-2729.2009.00313.x
  9. National Research Council, 2011, Successful K-12 STEM education: Identifying effective approaches in science, technology, engineering, and mathematics
  10. National Science Board, 2010, Preparing the next generation of STEM innovators: Identifying and developing our nation’s human capital
  11. Stark D. M., 2013, Ill-structured problems, scaffolding and problem-solving ability of novice nursing students
  12. Su Y., 2010, Journal of Educational Multimedia and Hypermedia, 19, 327
  13. Zhang J., 2000, Proceedings IFIP World Computer Congress, Educational Uses of Communication and Information Technologies, 344

Cited by

See all 346

Metadata

Authors
Belland, B. R., Walker, A. E., Kim, N. J., & Lefler, M.
Affiliations
Utah State University (US)
Venue
Review of Educational Research
Published
2016-10-11
Volume
87
Issue
2
Pages
309-344
ISSN
  • 0034-6543 (ISSN-L)
  • 1935-1046
OpenAlex
W2530735505
Cited by
346
References
127
Citations by year
2026: 62 · 2025: 44 · 2024: 35 · 2023: 46 · 2022: 52 · 2021: 31 · 2020: 29 · 2019: 22 · 2018: 11 · 2017: 9 · 2016: 5
FWCI
36.56
Type
journal article
Language
en
Open access
Locations
  • Review of Educational Research (publishedVersion, cc-by-nc)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
Topics
  • Social Sciences › Psychology › Developmental and Educational Psychology › Innovative Teaching and Learning Methods
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Visual and Cognitive Learning Processes
  • Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
Keywords
STEM education, cognitive outcomes, ill-structured problems, adult learners, primary education, educational technology effectiveness

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Synthesizing Results From Empirical Research on Computer-Based Scaffolding in STEM Education. https://app.cochat.ai/papers/10.3102/0034654316670999
The paper
Belland, B. R., Walker, A. E., Kim, N. J., & Lefler, M. (2017). Synthesizing Results From Empirical Research on Computer-Based Scaffolding in STEM Education. Review of Educational Research. https://doi.org/10.3102/0034654316670999

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.