Literature Searching, Source Appraisal and Synthesis

A literature search is not link collecting; it is a process for discovering what is known and how it is known

Literature serves three early functions: it helps avoid repeating a question already well addressed without justification, shows how constructs have previously been defined and measured, and helps distinguish stable findings from disagreement or genuine gaps. It does not mean searching until a paper confirms what we already believe. We therefore need a plan that can be explained: where we searched, which terms or concepts we used, what period was covered, and how relevance was judged.

When the goal is a systematic review, searching becomes a formal methodological component: planned information sources, search strategies, documentation and study-selection criteria are part of the method itself. The Cochrane Handbook describes systematic searching as an extensive and documented effort to identify eligible studies and make the search reproducible as far as possible. In this diploma, we borrow that discipline even for smaller literature reviews, while avoiding the label “systematic review” unless the requirements of that review type are actually followed.

Translate the question into search concepts, not one long sentence

Start with the concepts that cannot disappear from the question. In the completion example these might include digital learning, persistence or completion, and perhaps a specific factor such as video length if it becomes part of the question. List synonyms and related terms within each concept, using OR to broaden terms within a concept and AND to combine distinct concepts according to the database’s logic. Not every element of a research question must appear in the search. Some outcomes or comparisons may be poorly represented in titles or abstracts, and forcing them into the search can reduce sensitivity. Search design is iterative and should be tested against what it retrieves.

Source appraisal begins with relevance to the question, then considers authority, method and the degree of currency appropriate to the claim. An official source may be suitable for a current policy; an original study for an empirical estimate; a high-quality review for a body of studies; and an opinion piece may help reveal terminology or debate but cannot carry the same evidential weight as empirical research. “Newest is always best” is not a rule: a stable methodological concept may remain useful for years, while market statistics or changing policies need recent verification.

Move from separate summaries to synthesis

Synthesis does not stop at “Study A said this; Study B said that.” Organize evidence around a pattern, disagreement, or plausible explanation. For example, studies may measure persistence differently: last activity reached, unit completion, or intention to continue. Those results should not be combined as if they measure the same outcome. A good synthesis states where sources converge, where they diverge, and whether differences may reflect different populations, measurements or designs. This is how a defensible knowledge gap emerges.