Questions to Ask When Reading a Study or Informational Page

-Written by Kimberly Czotter, Feb 28, 2022.

Qualities of a "Good" Research Study

1. Has a sound, and clear research question (hypothesis).

Details exactly what the study attempts to answer. It helps the researchers stay on track and within their scope. Researchers often use the FINER criteria to outline their hypothesis:

F – feasible

I – interesting

N – novel

E – ethical

R – relevant

2. Follows a systematic approach or methodology.

Qualitative studies: collecting non-numerical data which may interpret subjects' experiences, behaviours, and beliefs. The methods include interviews, groups, and case studies. Can show correlational relationships.

Quantitative studies: the collection and analysis of numerical data which allows researchers to find causational relationships. These studies include surveys, experiments, and observations.

Mixed-methods studies: offer a combination of qualitative and quantitative study methods. This allows for a holistic interpretation of data.

3. Details the current understanding in the field.

By reviewing the current research, it will help you gain a more holistic understanding of the study and also understand if the study is pilot research or has already been studied. A literature review is the ideal type of study to read to gain a solid understanding of what is currently known and unknown.

4. Uses empirical, relevant data and analytical methods.

Empirical data is collected by researchers and is less biased and objective.

Quantitative data is analyzed via inferential and descriptive statistics which aids researchers in finding patterns, summarizing variables, and making predictions.

Qualitative data is analyzed by identifying themes and patterns from textual data.

5. Is generalizable and representative.

This is the ability for a sample to represent a larger community. Consequently, the participants' characteristics should match the target population as closely as possible. For example, with EDS, those affected are most often women so the research studies should contain a similar proportion of women (80-90%).

6. Logic guided.

This is important for spotting inconsistencies and fallacies in their findings and claims. Logical reasoning is important for making research meaningful.

7. External validity.

This means that the research findings can be applied to the real world. A study had high external validity if the data can be generalized to different situations, and applied broadly.

8. Reproducible, replicable, and transparent.

These are incredibly important aspects of research to ensure that the study findings are valid, credible, and trustworthy.

9. Acknowledges limitations and suggests future research.

Often these suggestions for future research stem from the limitations of the study and help identify what is missing in the database of research.

10. It is ethical!

Researchers should protect the dignity and rights of participants while the researchers act objective, accountable, and honest.

Look out for these "red flags":

11. Uses enough participants to be "valid."

A good general rule of thumb is that a quantitative study (ex. testing a medication) uses at least 50-100 participants to increase internal and external validity. If a study like this has only a few participants, one outlying result may lead to a significant change in the results. Most studies will determine an appropriate number of participants to reach "saturation" meaning that they can get statistically significant results from their study and that the results can be applicable and reliably interpreted.

In a qualitative study (ex. interviews) there should be at least 10-15 participants to get a good understanding of the major themes. Most often, preliminary (pilot) studies will be done on smaller groups in an interview style, to then create a mixed-methods approach. A mixed-method study uses both qualitative and quantitative data and showcases the strengths of both methods of study.

12. The presence of control groups.

Studies are strongest when there is a control group because you can see the actual effects of the intervention. Control groups may be with "matched" populations (ex. those without disease but have similar characteristics), within the population itself (ex. matched to other people with the same disease), or within the study participants (ex. measuring for a baseline and then after intervention). The strongest method is using the population itself.

-Written by Kimberly Czotter, Feb 28, 2022.

A review of study types, from "best" to "worst"

These studies evaluate all the available research in the specific area of interest. They discuss findings, patterns, and gaps in the research. They will combine the study results, and evaluate them statistically to determine if there is "enough" quality evidence to suggest a conclusion.

2. Randomized-Controlled Trial

This is an experimental type of study (quantitative) in which participants are randomly selected to receive either treatment or a placebo. This is most often used for medication trials. There are subtypes of the blind (participants don't know what they are getting) and double-blind variety (neither researcher nor participant know what treatment they are getting).

3. Quasi-experimental

This experimental study does NOT randomly assign participants to receive a placebo or treatment. This may be seen in studies where some patients cannot receive medication, or if it is impossible to "fake" a placebo. For example, people may be assigned to receive a medication when have reached a certain point of disease progression, whereas it may be unsafe for people who are more progressed.

There are three major subtypes: natural, nonequivalent, and regression discontinuity.

4. Cohort Study

This study type follows a group of people in order to understand their risk factors and outcomes. This is an observational study and is most famous for the Nurses' Health Study in where over 100,000 Nurses' health status has been followed for the last 40 years (LEARN MORE). These studies are useful in understanding disease progression, and potential risk factors - although, all results are correlational and not causational. Correlational research often can serve as the basis for researchers studying causational research.

5. Case-control Study

These studies often compare the history of a patient population to matched controls that do not have a condition. Ex.

6. Cross-sectional Survey

This observational study analyzes the data in sample populations during a specific time point. For example, this may look at the prevalence of women, men, and non-binary with a condition. These are helpful in understanding demographics, potential treatment options, etc. but are NOT causational nor experimental so often these are preliminary studies that prompt further research. They may come with potential bias if participants do not answer honestly (ex. saying they have a diagnosis when it has not been confirmed by a medical provider) which may skew data.

7. Case Reports

These are detailed histories of a small number of individual cases. Most often, these are done in a hospital and consist of 1-5 patients. These studies are helpful when analyzed in a meta-analysis OR in the case of a rare disease. Unfortunately, many of the studies on rare subtypes of EDS are case studies, as is the case with MALS.

-Written by Kimberly Czotter, Feb 28, 2022.