Page 63 - Secondary school students’ university readiness and their transition to university Els van Rooij

Chapter 3
2008, see Table 3.2). Table 3.1 Inclusion criteria
Inclusion criterion
1. e article has been peer-reviewed and published in an academic journal.
2. e data has been collected in 2000-2015.
3. e data has been collected in a Dutch or Flemish higher education institute.
4. e sample consisted of rst-year students.
5. e outcome variable was GPA, number of attained credits, or retention (or dropout).
6. e sample size was larger than 30.
7. e data was original.
8. e independent variables were within the scope of our review.
9. e article meets the quality criteria by AERA (2008).
Table 3.2 e eight principles of scienti c research as de ned by AERA (2008)
A. Development of a logical, evidence-based chain of reasoning
B. Methods appropriate to the questions posed
Number of articles excluded for not meeting this criterion
4 8 5 28 21
1 5 4 1
C. Observational or experimental designs and instruments that provide reliable and generalisable ndings
D. Data and analysis adequate to support ndings
E. Explication of procedures and results clearly and in detail, including speci cation of the population to which the ndings can be generalised
F. Adherence to professional norms of peer review
G. Dissemination of ndings to contribute to scienti c knowledge
H. Access to data for reanalysis, replication, and the opportunity to build on ndings*
*As we did not perform a meta-analysis, we did not enquire for each article whether the data would be available.
3.3.3 Initial and full-text screening, including quality assessment
From the list of hits from each database, we screened the articles on title and abstract. When the inclusion criteria were met in the abstract or when the abstract did not provide su cient information to decide whether or not the article met the criteria, the article would continue to the phase of full-text screening. In total, 133 articles survived the initial screening of title and abstract. Among these, there were 19 duplicates, so 114 articles were le to be screened completely. During full- text screening, the main question was whether the article met all inclusion criteria. e 114 articles were divided amongst the authors. To guard the reliability of screening, 15 articles were screened by two authors. Since in all cases the authors independently agreed on whether or not to include an article, we proceeded with having each article checked by one author. Full-text screening resulted in the exclusion of 77 articles, so only 37 studies met the inclusion criteria. During
62