Cited by 246Online Learning and AnalyticsStudent Assessment and FeedbackIntelligent Tutoring Systems and Adaptive Learning

The DOI matches its Crossref record; 137 of the 144 references it cites were found in a registry or an index, and 7 are not checked yet; none of the 115 sources its claims cite is checked for support yet.

A systematic review of the role of learning analytics in enhancing feedback practices in higher education

Open access
Banihashem, Seyyed Kazem, Noroozi, Omid, van Ginkel, Stan, Macfadyen, Leah P., Biemans, Harm J.A.·2022·Educational Research Review

Abstract

Learning analytics (LA) offers new opportunities to enrich feedback practices in higher education, but little is understood about the ways different LA can enhance feedback practices for educators and students. This systematic literature review maps the current state of implementation of LA to improve feedback practices in technology-mediated learning environments in higher education. We used strict inclusion criteria to select relevant studies that have investigated the role of LA on feedback practices. To identify common features of LA for feedback studies, we coded relevant publications using an analytical framework that identifies four key dimensions of LA systems: what (types of data), how (analytic methods), why (objectives), and how educators and students are served by LA (stakeholders). Based on findings, we propose a conceptual framework that can guide the implementation of LA for feedback systems and also suggest future empirical research in this area.

Checked 2026-10-10 · 7 references not checked yet

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

This systematic review maps 46 empirical studies on learning analytics (LA) for feedback in higher education using Chatti et al.’s what/how/why/who framework. It proposes an evidence-based conceptual framework in which stakeholder and objective should guide data collection and analytic method choices for LA-based feedback.

Key findings

  • LA feedback objectives most often targeted reflection (20 studies, 43%), personalization (13, 28%), prediction (11, 24%), assessment (11, 24%), adaptation (10, 22%), recommendation (10, 22%), and monitoring (6, 13%).
  • LA-supported feedback was linked to pedagogical outcomes including academic performance (17, 37%), engagement/community of inquiry (11, 24%), self-regulation (8, 17%), motivation/emotion (8, 17%), time management (3, 6%), perception of feedback (2, 4%), cognitive load (2, 4%), reduced procrastination (1, 2%), satisfaction (1, 2%), reflective thinking (1, 2%), and transactional distance (1, 2%).
  • Trace data was the dominant data type (23 codes, 36%), followed by assessment data (12, 18%), academic history data (11, 17%), and survey data (8, 12%); information visualization was the most common method (22 codes, 50%), then data mining (11, 25%), text analysis (8, 18%), and social network analysis (3, 7%).
  • Students were stakeholders in all included publications, while educators were mentioned in 15 of 46; the proposed framework orders decisions as stakeholder → objective → data → method, with analytics levels (descriptive, diagnostic, predictive, prescriptive) tied to different feedback services.

Methods

  • PRISMA systematic review; searched Web of Science, Scopus, ERIC, IEEE, and LAK proceedings with terms (“learning analytics” OR analytic* OR “data mining”) AND (feedback OR feedforward OR “formative assessment”); inclusion required English, 2011–2022, peer-reviewed journal articles, empirical studies, and higher education settings.
  • From 1,318 initial records, after removing duplicates/non-peer-reviewed and screening, 46 studies met criteria and passed quality appraisal using Theelen et al. (2019) criteria; all 46 met the minimum methodology score.
  • Publications were coded in ATLAS.ti 9 using a four-category scheme (what/how/why/who); inter-rater reliability on two randomly selected papers was 91% agreement (κ = 0.84, p < .001).

Limitations

  • Stated: limited to empirical studies, so relevant reviews/conceptual papers may be missed; selected databases may miss unindexed studies; publication bias likely; findings are not generalizable to K-12; the review did not identify gaps in feedback features/purposes or investigate effect sizes of different data types.
  • Stated: no distinction was made between educator feedback and peer feedback, and no analysis of how LA may affect affective, cognitive, and constructive features of feedback quality.
  • Evident from the text: the coding scheme in Table 3 assigns the same code “C5” to both Personalization and Recommendation, and some table reference lists appear truncated or misaligned; these presentation issues may affect transparency of coding/reporting.

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

Integrity signals

Registry

Crossref · 10.1016/j.edurev.2022.100489

Cites

137 of 144 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 of 115 cited sources checked · 0 supported · 0 flagged

Show all 91 claims

  • Probably not a substantial claim. 57
  • It may state two separate claims as one. 39
  • The quote only partly states this claim. 3
89 of 89
  • The paper relies on this without support.

    Only few review studies have focused on providing an overview of the role of LA in feedback.

    “However, only few review studies have focused on providing an overview of the role of LA in feedback.”
  • The paper relies on this without support.

    According to the literature, there is no clear understanding of how LA tools can support feedback practices and how LA's use for feedback purposes in higher education can be guided.

    “According to the literature, there is no clear understanding of how LA tools can support feedback practices and how LA ' s use for feedback purposes in higher education can be guided.”
  • The paper relies on this without support.

    To date, no conceptual frameworks have been offered to guide such role of LA for the design of feedback practices.

    “To date, however, no conceptual frameworks have been offered to guide such role of LA for the design of feedback practices.”
  • The paper relies on this without support.

    There is no overview at which levels social network analysis methods can be used.

    “For example, there is no overview at which levels social network analysis methods can be used.”
  • The paper relies on this without support.

    For feedback practices in higher education, educators and students tend to be the key stakeholders.

    “It might be clear that for feedback practices in higher education, educators and students tend to be the key stakeholders”
  • The paper relies on this without support.

    Little is known about how educators and students have benefitted from LA applications in higher education.

    “however, little is known about how educators and students have benefitted from LA applications in higher education?”
  • The paper relies on this without support.

    Survey data usually contain high validity because they have already been validated by other scholars.

    “What is important to note about the survey data is that this type of data usually contain a high validity, since they have already been validated by other scholars.”
  • The authors state this as an assumption.

    The selected set of literature databases covers the most relevant publications.

    “while we feel that our selected set of literature databases covers the most relevant publications”
8 of 8

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?

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.

  • UK study percentage is too high

    Minor

    Three of 46 studies is about 6.5%, which rounds to 7%, not 8%.

    Reported 8 · recomputed 6.521739

  • Conference papers included despite a journals-only criterion

    Minor

    The inclusion criteria restrict publications to scholarly journal articles, but the final set includes six conference-proceedings papers.

2 of 2

Citations

Cites

Flagged first, then by citations

Show all 138 references

Cited by

Sorted by citations

Items 201–245 of 245

Metadata

Authors
Banihashem, S. K., Noroozi, O., van Ginkel, S., Macfadyen, L. P., & Biemans, H. J. A.
Affiliations
  • Open University of the Netherlands (NL)
  • Wageningen University & Research (NL)
  • University of British Columbia (CA)
Corresponding author
Banihashem, Seyyed Kazem
Venue
Educational Research Review
Publisher
Elsevier BV
Published
2022-11
Volume
37
Pages
100489
ISSN
1747-938X (ISSN-L)
OpenAlex
W4308066047
Cited by
246
References
138
Citations by year
2026: 74 · 2025: 74 · 2024: 57 · 2023: 40 · 2022: 1
FWCI
7.05
Type
journal article
Language
en
Open access
Locations
  • Educational Research Review (publishedVersion, cc-by)
  • Socio-Environmental Systems Modeling (submittedVersion, cc-by)
  • DSpace (Open University in the Netherlands) (submittedVersion, cc-by)
  • Data Archiving and Networked Services (DANS) (submittedVersion, cc-by)
Topics
  • Physical Sciences › Computer Science › Computer Science Applications › Online Learning and Analytics
  • Social Sciences › Education › Student Assessment and Feedback
  • Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
Keywords
learning analytics, feedback practices, higher education, technology-mediated learning environments, conceptual framework, student feedback, analytical methods, stakeholders

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.

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

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

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