Lupus research is critical for developing new treatments, diagnostics, and wellness strategies.
There are often more questions than answers when it comes to lupus research. It can be quite frustrating to see contradictory information about treatments or lifestyle changes.
After reading enough articles on the internet about lupus research, you may even find yourself asking: What do we actually know about lupus?
There are many challenges with sharing research information. But, looking out for these 3 common issues will help you better understand the value of new research (and identify the shortcomings!).
- Oversimplification of results
- Studies often measure a number of variables or factors, some of which may be changed by the study and some that may not. Statistically significant findings make it into abstracts (and to the news), but may not be contextualized properly
- Extrapolation is a specific type of estimating that is not reflected in measurable data. Rather, estimates are made based on other measurements
- There are occasionally efforts to extrapolate information about humans from research conducted on animals. Here is an article that speaks to this specific challenge
- Ethical and logistical challenges with conducting studies in the real world
- People are different and not all variabilities can be measured. This can cause for unknown or unmeasurable factors to influence results
Underlying all of these issues are statistical challenges. Nature published a piece in 2014 that highlighted problems with P values (a measure of statistic power).
Types of lupus research
Scientists are typically cautious to make grand claims about their research; journalists and bloggers can aim for excitement and optimism. This is (partially) why reading health journals is bland while reading the health section of your favorite news site can be fascinating.
(NOTE: At LupusCorner, we work to combat this by contextualizing new results with information from prior studies. In this way, we can share the excitement about new findings while providing an overview of existing research.)
One of the reasons that scientists can be less inclined to make grand claims about their research is that certain study designs do not allow for grand claims to be made.
When it comes to research design, there are lots of terms and specifics that can make it difficult to understand the differences — from qualitative/quantitative to fixed/flexible designs. Below, we will look briefly at 3 design types. However, here is additional information about study design.
Descriptive and correlational studies
Power of results: Low
Types: case studies; observational studies, surveys
Useful in research to:
- provide insight into how things occur in the real world
- help form hypotheses for experiments
- provide feedback on clinical treatments/practices
Case studies often occur on very small samples (a case study can even be written about a single participant or case). Often, they highlight phenomena not noted previously. In this way, they may trigger subsequent research.
Observational studies are used with the independent variable of a study is beyond the control of researchers. This can be true because of ethical concerns. For example, not all people want the same treatments options/surgeries and researchers cannot make research participants receive treatments.
These types of studies often result in correlational results. Correlation is not causation (though this is a common mistake made when interpreting results of this type).
Power of results: High
Types: Controlled experiments with randomization (Clinical Trials)
Useful in research to:
- prove a hypothesis by comparing identical groups
- control confounding factors
Clinical trials are a complicated process that are broken into phases. To learn more about clinical trials, click here.
The goal of experiments is to learn something about an intervention by attempting to control for all other variables. In this instance, an intervention is done to, or given to, one group but not to the other. Then, certain factors are measured and, using statistics, differences are measured.
Multiple study analysis
Power of results: Very high
Types: systematic reviews; literature reviews; meta-analysis
Useful in research to: Analyze a large number of studies in order to make larger or more profound claims.
By reviewing or analyzing a large number of studies together, more advanced claims can be made regarding the results. In a review, researchers will look for all the studies relating to a specific topic. Then, they synthesize the results and look for patterns and results that emerge across the studies.
A meta-analysis is similar to a review, but it goes even further by combining the results of many scientific studies using statistical analysis. Essentially, a meta-analysis attempts to reduce the variabilities in the studies by looking at a number of studies together, statistically.
Using research design information
Understanding research design will help you better evaluate the research that you read. Reading the Method section of research papers can be tedious. But, by reading and listening for key words (e.g., case study, observational study, experiment, review, meta-analysis), you can quickly perform your own analysis of the study design.
If the claims or results seem to extend beyond the power of the study design, it’s possible that the author is making 1 of the common mistakes mentioned above.