OE Research Guide

What is a Traditional Literature Review?

James Neill
Last updated:
08 May 2006

Also see: How to write a literature review (class notes for graduate students)

A "traditional" literature review provides an overview of the research findings on particular topics.  A traditional literature is written by examining a body of published work, then writing a critical summary (an impressionistic overview) of the body of literature.  The purpose of a literature review is make clear for a reader what the research collectively indicates with regard to a particular issue or question.

Literature reviews are vital documents for organizing and making accessible the major findings in an area of inquiry.  Literature reviews are often used, for example, to inform policy and future research directions.  Conducting reviews of literature is particularly important as a field of inquiry expands or changes.

A good traditional literature review will usually consider research evidence in light of an established theoretical framework.  A theoretical context can enable meaningful synthesis and interpretation of research findings (e.g., see the review by Barret & Greenaway, 1995b). 

A traditional literature review is not foolproof - it has some strengths and weaknesses.  The method involves authors reading original (primary) studies , taking notes, organizing themes, and immers ing themselves in the research literature in order to develop a comprehensive understanding of the main findings (also see How to Write a Literature Review) .  R eviews of literature should be subject to peer-review to help ensure that all relevant literature is considered and that the approach is systematic and as valid as possible.

  • Strength: W hen a traditional review is systematically conducted by an expert in the field, insightful, valid syntheses of the research literature can be developed and sometimes unique insight is generated .
  • Weakness: T he traditional literature method leave s authors vulnerable to unintentional and intentional bias in the selection, interpretation and organization of content.  In addition, it is difficult for the statistical power of empirical studies to be considered in interpreting many studies.  As a result traditional reviewers regularly emphasize the statistical significance results , rather than the effect sizes.  In fields with research that has low power, traditional literature reviews may be compromised by this potentially serious issue[1].  The meta-analytic method of literature reviewing is one way in which this issue can be overcome.

Footnote

[1] When the power of research studies within a field of inquiry is systematically high or low, then the meaningfulness and validity of the reported results may be questionable (Lipsey & Wilson, 1993). In the case of reviewing outdoor education research literature, the concern is to do with the power of the research being too low.  Lipsey and Wilson (1993, p. 1200) explain that “when alpha is set to the usual levels (e.g., .05) to limit Type I error, Type II error will be unrestrained and can range very high (e.g., 50%-90%) unless sample sizes are quite large (Schmidt, 1992)…Because…the null hypothesis is generally false in…[psychological] treatment research…and, also, sample sizes are modest, high Type II error rates will result in a larger proportion of spurious null (statistically nonsignificant) results in treatment research”.  In Lipsey and Wilson’s study of psychological treatment effects, they used 156 meta-analyses, with an average sample size of 134, or 67 each in the treatment and comparison groups.  The statistical power for that sample size, with alpha equal to .05 and a treatment effect of 0.47, was 0.76.  Despite the positive treatment effect in this average case, 24% of studies would be expected to yield statistically nonsignificant results.  When the ES is below .47, or sample size falls below 67, the power drops off quite sharply.  With ES = .20 and n = 50, not an untypical situation in outdoor education research, the Type II error rate will be a disturbing 83% (Lipsey & Wilson, 1993).  Thus, for four out of every five outdoor education research studies which find report non-significant results, there may, in fact, have been positive effects of interest.