- Reporting quantitative results (8 minute read time)
Content warning: Brief discussion of violence against women.
16.1 Reporting quantitative results
Learners will be able to…
- Execute a quantitative research report using key elements for accuracy and openness
So you’ve completed your quantitative analyses and are ready to report your results. We’re going to spend some time talking about what matters in quantitative research reports, but the very first thing to understand is this: openness with your data and analyses is key. You should never hide what you did to get to a particular conclusion and, if someone wanted to and could ethically access your data, they should be able to replicate more or less exactly what you did. While your quantitative report won’t have every single step you took to get to your conclusion, it should have plenty of detail so someone can get the picture.
Below, I’m going to take you through the key elements of a quantitative research report. This overview is pretty general and conceptual, and it will be helpful for you to look at existing scholarly articles that deal with quantitative research (like ones in your literature review) to see the structure applied. Also keep in mind that your instructor may want the sections broken out slightly differently; nonetheless, the content I outline below should be in your research report.
Introduction and literature review
These are what you’re working on building with your research proposal this semester. They should be included as part of your research report so that readers have enough information to evaluate your research for themselves. What’s here should be very similar to the introduction and literature review from your research proposal, where you described the literature relevant to the study you wanted to do. With your results in hand, though, you may find that you have to add information to the literature you wrote previously to help orient the reader of the report to important topics needed to understand the results of your study.
In this section, you should explicitly lay out your study design – for instance, if it was experimental, be specific about the type of experimental design. Discuss the type of sampling that you used, if that’s applicable to your project. You should also go into a general description of your data, including the time period, any exclusions you made from the original data set and the source – i.e., did you collect it yourself or was it secondary data? Next, talk about the specific statistical methods you used, like t-tests, Chi-square tests, or regression analyses. For descriptive statistics, you can be relatively general – you don’t need to say “I looked at means and medians,” for instance. You need to provide enough information here that someone could replicate what you did.
In this section, you should also discuss how you your variables. What did you mean when you asked about educational attainment – did you ask for a grade number, or did you ask them to pick a range that you turned into a category? This is key information for readers to understand your research. Remember when you were looking for ways to operationalize your variables? Be the kind of author who provides enough information on operationalization so people can actually understand what they did.
You’re going to run lots of different analyses to settle on what finally makes sense to get a result – positive or negative – for your study. For this section, you’re going to provide tables with descriptions of your sample, including, but not limited to, sample size, frequencies of sample characteristics like race and gender, levels of measurement, appropriate measures of central tendency, standard deviations and variances. Here you will also want to focus on the analyses you used to actually draw whatever conclusion you settled on, both descriptive and inferential (i.e., bivariate or multivariate).
The actual statistics you report depend entirely on the kind of statistical analysis you do. For instance, if you’re reporting on a logistic regression, it’s going to look a little different than reporting on an ANOVA. In the previous chapter, we provided links to open textbooks that detail how to conduct quantitative data analysis. You should look at these resources and consult with your research professor to help you determine what is expected in a report about the particular statistical method you used.
The important thing to remember here – as we mentioned above – is that you need to be totally transparent about your results, even and especially if they don’t support your hypothesis. There is value in a disproved hypothesis, too – you now know something about how the state of the world is not.
In this section, you’re going to connect your statistical results back to your hypothesis and discuss whether your results support your hypothesis or not. You are also going to talk about what the results mean for the larger field of study of which your research is a part, the implications of your findings if you’re evaluating some kind of intervention, and how your research relates to what is already out there in this field. When your research doesn’t pan out the way you expect, if you’re able to make some educated guesses as to why this might be (supported by literature if possible, but practice wisdom works too), share those as well.
Let’s take a minute to talk about what happens when your findings disprove your hypothesis or actually indicate something negative about the group you are studying. The discussion section is where you can contextualize “negative” findings. For example, say you conducted a study that indicated that a certain group is more likely to commit violent crime. Here, you have an opportunity to talk about why this might be the case outside of their membership in that group, and how membership in that group does not automatically mean someone will commit a violent crime. You can present mitigating factors, like a history of personal and community trauma. It’s extremely important to provide this relevant context so that your results are more difficult to use against a group you are studying in a way that doesn’t reflect your actual findings.
In this section, you’re going to critique your own study. What are the advantages, disadvantages, and trade-offs of what you did to define and analyze your variables? Some questions you might consider include: What limits the study’s applicability to the population at large? Were there trade-offs you had to make between rigor and available data? Did the statistical analyses you used mean that you could only get certain types of results? What would have made the study more widely applicable or more useful for a certain group? You should be thinking about this throughout the analysis process so you can properly contextualize your results.
In this section, you may also consider discussing any threats to that you identified and whether you think you can generalize your research. Finally, if you used any measurement tools that haven’t been validated yet, discuss how this could have affected your results.
Significance and conclusions
Finally, you want to use the conclusions section to bring it full circle for your reader – why did this research matter? Talk about how it contributed to knowledge around the topic and how might it be used to further practice. Identify and discuss ethical implications of your findings for social workers and social work research. Finally, make sure to talk about the next steps for you, other researchers, or policy-makers based on your research findings.
- Your quantitative research report should provide the reader with transparent, replicable methods and put your research into the context of existing literature, real-world practice and social work ethics.
- Think about the research project you are building now. What could a negative finding be, and how might you provide your reader with context to ensure that you are not harming your study population?
The process of determining how to measure a construct that cannot be directly observed
Ability to say that one variable "causes" something to happen to another variable. Very important to assess when thinking about studies that examine causation such as experimental or quasi-experimental designs.