A Meta-Analysis of Educational Data Mining on Improvements in Learning Outcomes.
Bibliographic record
Abstract
A meta-synthesis study was conducted of 60 research studies on educational data mining (EDM) and their impacts on and outcomes for improving learning outcomes. After an overview, an examination of these outcomes is provided (Romero, Ventura, Espejo, & Hervas, 2008; Romero, et al., 2011). Then, a review of other EDM-related research published after 2008 (88 studies) was completed. Thirty-nine of those studies also offered an overview of EDM's impact on learning outcomes. In addition, 12 of the 39 studies investigated the efficacy of EDM for learning outcomes. EDM characteristics (i.e., tools, techniques, models, procedures, measures, and results) were examined in each of the 12 studies. Ninety-four of the total 148 studies showed positive results for EDM. Directions for future research are discussed. Keywords: Educational data mining, learning outcomes, prediction, classification Introduction Recently, the concept of educational data mining (EDM) has witnessed dramatic worldwide growth in the field of education. EDM has gained increased attention as a process that provides useful data necessary for decision-making in education institutions (Kusiak, 2002). Specifically, teacher education programs are recognizing EDM as a useful analytical tool that may lead to improvements in learning outcomes. Also, research (i.e., Al-Shammari, 2011; Ogundokun, 2011) has been placing great emphasis on the improvement of learning outcomes. Several research studies (e.g., Barros & Verdejo, 2000; Chapman & Bloxham, 2004; Devine, Hossain, Harvey & Baur, 2011; Minaei-Bidgoli, Kortmeyer & Punch, 2004; Ranjan & Ranjan, 2010; Thai-Nghe, Drumond, Horvath & Schmidt-Thieme, 2011; Xiong, Pardos & Heffeman, 2011; Yudelson, Medvedeva, Legowski, Castine, Jukic & Crowley, 2006) have investigated the effects of EDM on learning outcomes. Other research (Delavari, Beikzadeh, & Phon-Ammuaisuk, 2008) has cited EDM as a useful tool in improving learning outcomes due to its ability to identify at-risk students and predict their future performance in learning settings. Sparks (2011) found that EDM provides answers to questions relating to student performance. Campbell, DeBlois, and Oblinger (2007) also stated that EDM can answer the call for accountability through academic analytics, which is emerging as a new tool for a new era (p. 40). This research has three purposes: to provide an overview of EDM; to offer a brief overview of research on EDM; and to investigate and then discuss how the use of EDM improves learning outcomes. Overview of EDM EDM is a new tool used in education to uncover useful information in or hidden relationships among large amounts of electronic data stored in a school's system (Baker, 2011). EDM uses reliable techniques in multiple analytic procedures that ultimately reveal hidden information that helps teachers, administrators, and others in improving learning outcomes (Ogor, 2007). According to Romero and Ventura (2010), EDM answers questions related to what a student actually knows and whether a student is learning. EDM works much like Learning Analytics, except that EDM basically addresses the development of new methods of making discoveries for data analysis while Learning Analytics addresses only the application of known methods (Baker, 2011). EDM methods differ from other methods used in general data mining due to the complexity of interrelationships among types of data and data sets. For example, there are multiple levels of hierarchy in educational data (Baker, 2011; Campbell, DeBlois & Oblinger, 2007; Romero & Ventura, 2010). EDM borrows many of its applications from machine learning and artificial intelligence (Baker, 2011). By the late 1990s, data mining had split off from artificial intelligence to stand on its own (Romero, Ventura, Espejo, & Hervas, 2008). In 2008, the field of education officially witnessed the growth of EDM at the First Annual Conference on Educational Data Mining, which was held in Montreal, Quebec, Canada in June 20-21, 2008 (Winters, 2006). …
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.043 | 0.112 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.012 | 0.040 |
| Bibliometrics | 0.013 | 0.009 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".