Are you struggling to find the best way to write a literature review?
We have compiled some of the most important points in this blog post to help you get started. From defining what a literature review is, to how it can be formatted and written, we cover it all.
This guide will help you write an exceptional literature review that thoroughly examines the current state of research on your topic, considers how other scholars have approached it, and concludes with your own recommendations for future study and research.
Literature Review Definition
Literature reviews are an important part of most fields of study. Many times it’s not enough to simply present all available research on a topic as it is.
Instead, research must be evaluated by scholars in that field who can critically assess previous studies, decide which ones are most important and relevant to their work, and explain how they integrate that information into their own study or research.
In short, writing a literature review is a way for them to introduce themselves to their new colleagues in a scholarly context.
The best literature reviews are objective and thorough.
They summarize key aspects of the original study while also pointing out limitations or flaws in the design.
While reviewing the original study, keep these initial questions in mind:
- What was studied?
- How was it studied?
- What were the findings?
- What conclusions did they draw from those findings? What major questions remain unanswered?
- Is there anything else worth noting about the study that might affect its interpretation or applicability to your research question?
Why Write a Literature Review
Whether you are a student or an academic, you may need to write a literature review at some point in your career.
Students use literature reviews to explore potential topics and methods of analysis.
Academics use them to evaluate past research, share thoughts on where the field should go next, and highlight gaps in knowledge so they can direct future studies.
Your college professors, supervisors, and editors want to know that you’ve spent time examining your sources carefully and critically; not just throwing together facts without thought.
A literature review is your chance to show that you’re a careful researcher who isn’t afraid to be honest about what other scholars have discovered.
It’s a chance to show off your critical thinking skills. In short, it’s a way for you to demonstrate what makes you different from other researchers—and how you can contribute new knowledge to your field.
How to Write a Literature Review
Step 1: Define the scope of your literature review.
You’ll want to start by defining the scope in terms of keywords that relate to your topic.
For example, if you’re interested in investigating how social media impacts mental health, you might include keywords like social media, mental health, internet addiction, etc.
If you’re looking to compare results from three experimental treatments (i.e., diet, exercise, medication), then you might focus on words like diet, exercise, medication, and treatment outcome.
Keep in mind that as you read more and more articles related to your topic (that will provide useful evidence for your paper), the scope may change based on what specific aspects of the topic interest you most.
As you do your research, don’t forget to revisit the goal of your review.
For the sake of simplicity, you can create a list of your topic-related keywords in the following format: Keyword1, Keyword2, and so on.
What you want to do is rank your keywords in order of importance, starting with the one that’s most important to your topic and working down to the least.
This will give you a rough idea of the organization you’ll follow. Be sure to put your main keyword in bold and others that you plan on mentioning in italics.
Consider our example: here is a possible list of keywords to start with: social media, mental health, internet addiction, comparative research, and treatment outcome.
When you find a research article that looks interesting and relevant to your topic, search the abstract and look for keywords that are in line with your list.
You should also check the bibliography and the references listed in the article to see if they are in line with your keywords.
Make a note of any keywords that you haven’t yet included, and add them to your list.
Once you have collected several articles, return to your list and group them by their relevance to each keyword.
Based on the keywords, you can now divide your list into four categories: Social Media and Mental Health, Internet Addiction, Comparative Research, and Treatment Outcome.
List the items in each category alphabetically or numerically to organize them.
At this point, you might want to revise your list of keywords. What keywords are the most relevant to your topic?
These would be the ones that you’d want to include in your review.
Note that this is a constantly evolving process and you can keep adding or removing keywords as you read through more literature.
Step 2: Choose a form for your literature review.
The next step is to choose the form of your literature review.
The most common forms are traditional or narrative, systematic, meta-synthesis, and comparative research.
Traditional or narrative Literature Review
This type of review tells a story about the development of ideas relating to your topic over time.
It generally starts with an introduction to the subject matter, followed by an explanation of existing theories and research, leading up to new conclusions drawn based on findings.
Unlike other forms such as a systematic review, it doesn’t systematically identify potential studies.
Rather, it reviews what has been published in recent journals.
A disadvantage of this type of review is that it often includes subjective observations and personal opinion.
This may make the reviewer seem biased, especially when interpreting their conclusions without presenting alternative interpretations or giving a rationale for why certain points were excluded from the analysis.
One way to overcome this bias is to use a mixed methods approach.
This method employs qualitative and quantitative research methods in tandem, which helps the researcher understand patterns in data and provide insight on the decision-making process.
For example, you could ask participants to reflect on why they feel something is true, or you could use qualitative interviews with people who are experts in your field of study.
Systematic Literature Review
Systematic reviews are systematic and objective.
They follow a rigorous methodology for searching for, assessing, appraising, and presenting evidence-based research on a specific topic.
You might also see these referred to as critical reviews of research.
To do a systematic review, researchers first define a problem, question, or hypothesis to examine.
Then they conduct searches of databases and other sources looking for all appropriate studies that answer the problem question or hypothesis.
Studies found must meet predetermined inclusion criteria (e.g., related to the original issue).
After retrieving all eligible studies, quality ratings are given and adverse effects assessed for those eligible studies meeting all inclusion criteria.
Finally, the research is summarized and the findings presented.
Systematic reviews are a great starting point for someone who wants to learn about the current state of research on a topic. It’s important to note that systematic reviews will not necessarily produce strong evidence for your own hypotheses.
That’s because systematic reviews require pre-specified questions or topics, and it’s likely that there are many aspects of your study not covered in the review.
This type of review provides detailed analyses of past research that can be used to draw inferences about future directions for research, rather than just describing past work.
It goes beyond summarizing results by synthesizing information from multiple studies.
By combining statistical tests, meta-analysis quantitatively synthesizes empirical evidence across different experiments/studies in order to provide stronger causal relationships between variables than provided by any individual study alone.
Essentially, it uses a measure of effect size to compute the degree to which we can trust the estimate of a particular parameter.
This means that if we have five studies, each with 100 participants, and the four studies show a correlation of .5 and the fifth shows a correlation of .3, then the average correlation would be .4.
While it’s possible for individual studies to show a high degree of variability in their estimates, meta-analysis reduces this variation.
It does this by using measures of effect size, like correlation coefficients.
There are various types of meta-analyses, including fixed and random effects models.
Fixed models assume that the underlying distribution is the same for all studies, while random models allow for heterogeneity in study samples.
In other words, the fixed model assumes that all studies are estimating the same thing and that one study is as good as another.
In contrast, the random model allows for differences in study samples, and so it is a more realistic model.
Generally speaking, a meta-analysis with a fixed effects model is more conservative than a random effects model because it treats all studies equally.
However, this is not always desirable or necessary. If you’re interested in how two things correlate within the same sample, you might want to use a random effects model instead.
What’s more, if you want to compare two treatments and calculate an effect size statistic but find that both treatments were tested in the same experiment, then you’ll need to use a fixed effects model.
This is a type of literature review that is quantitative in nature and compares the efficacy of at least two interventions.
Examples include a review of the relative effectiveness of drug A versus drug B, comparing diet A to diet B, and comparing treatment 1 to treatment 2.
These reviews typically employ descriptive statistics to summarize the data and are statistically more sophisticated than traditional reviews.
For example, a comparative review might employ correlation analysis to assess the relationship between two independent variables.
This is accomplished by calculating the strength of the association between two phenomena.
The strength of association is measured with a correlation coefficient, which ranges from -1 to +1.
Values close to -1 indicate that two phenomena are inversely related, whereas values close to +1 indicate that two phenomena are directly related. The closer the value is to 0, the weaker the association.
Step 3: Find Studies to Review
Once you’ve determined the best design for your review, it’s time to start looking for articles that fit the criteria.
Depending on the specific research question, and what methodology you are most interested in (systematic reviews, meta-synthesis, comparative research), some databases may be better suited for finding relevant articles.
Systematic reviews, for instance, are usually published in a peer-reviewed journal and focus exclusively on randomized controlled trials.
These databases tend to be indexed for medical subject headings (MeSH) and free-text terms.
Other databases such as PsycINFO index a broader range of sources and contain mostly peer-reviewed journal articles.
Sources in this database are often indexed for keywords, subjects, and author names.
Another database that is useful for literature reviews is Google Scholar.
Google Scholar indexes a variety of journals and databases including Medline, Scopus, and Web of Science.
Free access to these articles may be available as soon as they are published online; however, you will still need to pay for access to the full text.
If you’re conducting a systematic review, make sure to identify any eligibility criteria before starting your search.
If you’re conducting a meta-analysis or comparative research study, do the same thing for inclusion/exclusion criteria.
Compile a list of potential studies and determine if they meet your eligibility criteria.
You can use filters in databases like PubMed to find all articles about, say, obesity or diabetes that were published within the last 10 years.
Step 4: Review your Results
After selecting studies that fit your criteria, it’s time to dig into them.
Some of these studies will be more relevant than others based on how well they answer your research question.
But regardless of relevance level or quality of methodology, you need to assess whether or not a study is a good fit for your meta-analysis.
This will involve systematically reviewing each study’s design, participants, methods used in collecting data and analyzing it (i.e., statistical tests), results (including confidence intervals), and limitations.
To begin, check the abstract and summary to see if the study meets your eligibility criteria.
This is also a good place to see the major limitations of the study, especially in case you want to avoid studies that are underpowered, lack external validity, or have large variability in participant characteristics.
If a given article seems to meet your criteria, go ahead and read it thoroughly.
Again, note limitations when you come across them. If a study does not provide enough information for you to assess it, move on.
If the article is in a language other than English, you will need to decide whether or not it will be worth your time to translate it.
Sometimes the title and abstract are enough to give you an idea of the article’s content and limitations.
Ultimately, this process boils down to asking yourself two questions: Does this study meet my criteria? And is this study high quality?
Don’t forget to ask yourself these questions after reading every single one of the studies you identified in step 3.
Step 5: Synthesizing Results
In order to synthesize results, you will want to organize studies in a way that makes sense to you.
If there is more than one study analyzing similar data (e.g., different interventions or patient populations), combine them into one table or graph.
This will allow you to more easily compare the findings.
If a study doesn’t seem to fit, consider using it as a supplement for a related topic that is mentioned in your review.
For example, if you are discussing the efficacy of cognitive behavioral therapy (CBT) for patients who suffer from depression, and you came across a study on CBT for patients who suffer from bipolar disorder but did not suffer from depression, consider adding this article as a supplement or appendix to your review.
If a study includes analyses for more than one type of outcome measure, make sure to include all outcomes in your synthesis tables or graphs.
Remember that your conclusions should be generalizable to any type of reader – for example, clinical practitioners, researchers, policymakers – so try to present findings from all perspectives whenever possible.
Most papers take about 20 minutes to summarize, which might sound like a lot until you think about all the work you’ve done up until this point.
Keep in mind that just because a paper isn’t a perfect fit for your paper doesn’t mean it isn’t worth mentioning at all; sometimes including supplemental studies can create new insights or lead to new directions for future research.
Draft your results section by summarizing your findings.
Be concise and clear, while highlighting the most important findings.
If a study is not a great fit for your review, summarize it briefly and mention the most important limitations. If a study has multiple limitations, list the most significant ones first.
Step 6: Analyze the Information Gathered
The discussion section of your review is where you share your thoughts on the strengths and limitations of the current body of evidence.
It is here that you will identify gaps in the current body of knowledge and propose ideas for future research.
Your discussion section should include a critical analysis of your results and their implications for practice, policy, and future research.
For example, let’s say that one of your major findings was that smoking cessation counseling is effective for pregnant women who smoke during pregnancy.
Your discussion should include a critical analysis of your results and their implications for practice, policy, and future research.
For example, let’s say that one of your major findings was that smoking cessation counseling is effective for pregnant women who smoke during pregnancy.
Your discussion could then focus on how this finding could be used to improve public health initiatives targeting pregnant smokers by focusing resources on providing cessation counseling rather than other measures such as simply educating them about the dangers of smoking or asking them to quit cold turkey without support.
You could also highlight the fact that pregnant smokers often have difficulty quitting even when they know that smoking harms their fetus, so this intervention may have additional benefits.
You might also discuss whether the intervention would have been equally effective had it been conducted in a group setting rather than individually, or whether it would have been helpful to incorporate feedback from another person (such as an obstetrician).
Another idea would be to explore why some participants were able to stop smoking despite receiving only counseling and not nicotine replacement therapy.
Perhaps it is due to the counselor-patient relationship, and a study exploring this possibility would be beneficial.
Ultimately, you want to ensure that your readers will be able to make a change in practice, policy, or future research based on the findings of your review. If you cannot find a study on a particular topic, it is always appropriate to suggest one as a future research area.
Step 7: Provide Recommendations and Conclude
In the last step, you’ll need to synthesize all of these points together into a coherent conclusion before finishing up your paper.
Remember, the conclusion does not restate what has already been saying.
Rather, it summarizes the key points made throughout the paper.
Make sure that your conclusions are supported by previous parts of the text; any recommendations should follow logically from what has already been discussed.
Avoid saying In conclusion at the beginning of each paragraph – instead, summarize each paragraph within its own sentence before moving on to a new point.
Refer back to the aim of your review if you’re struggling to remember what you’ve already covered.
Make sure that the introduction and abstract match each other.
Finally, remember that it is important to provide practical recommendations for both practitioners and policymakers alike.
For example, if you found that more attention needs to be paid to reducing the risk of developing heart disease among postmenopausal women through physical activity programs, you can recommend specific types of exercise and how much time needs to be spent exercising per week.
Similarly, if your review shows smoking cessation counseling to be ineffective for pregnant women, you can recommend that future interventions utilize a different form of counseling.
Regardless of the type of review, it is important to come to a strong conclusion that offers real and actionable advice for those in the field.