For beginning learners in statistics, we always wonder what observational studies are, the difference between observational studies and randomized control trials, and the types of observational studies. This article will discuss these above topics in detail and provide examples of each type of observational study.
In the observational studies, researchers observe the effect of risk factors, treatment, or other interventions without control over the exposure. It usually leads association between exposure and outcome. However, in the randomized control trial, participants are randomly separated into the control and treatment groups. Researchers can control exposure and find the causal relationships between the response variable and independent variables. The observational studies have different types, such as survey/cross-sectional study, cohort study, and case-control study, which I will discuss further.
Cross Sectional Study
In the cross-sectional study, we split the sample population into two groups, exposed and unexposed, and then observed whether the participants in two groups have the outcome. An example of a cross-sectional study can be present as follows: The researcher split participants into different ages, ethnicities, geographical locations, and social backgrounds. Then to observe whether or not the participants in different groups are more prone to have cancer. If a significant number of participants from particular geographical locations are more prone to have cancer, the particular geographical locations may be associated with more probability to have cancer.
However, some limitations are regarding the cross-sectional study. If the sample population is not representative, which means not a large and heterogeneous study population, it will have selection bias. It will lead to the wrong conclusion drawn from the results.
Cohort Study
In the cohort study, we split the sample population into two groups exposed to the intervention and unexposed to an intervention. After a while, observe whether or not the participants in the two groups have the outcome. A cohort study can be present as follows: To focus on how sun exposure damages the skin’s general health. Group A, which includes 20 participants use UVA-blocking sunscreen. Group B consists of 20 participants who do not use any UVA-blocking sunscreen. After a month’s tournament, participants’ skin in both groups was analyzed for texture, sun damage, and burns. The researchers compare skin damage in both groups based on the use of sunscreen.
However, there are some limitations regarding cohort study. It usually takes months or years to see significant results and expensive to conduct a cohort study. Therefore, it may not be ideal for researchers to analyze results in a short time.Case-Control Study
Unlike cohort study or cross-sectional study, a case-control study first split participants into two groups based on whether they have the outcome and then check if participants have been exposed to intervention before having an outcome or not having outcome.
An example of a case-control study can be present as follows: We are focus on how a case-control study provides an example of breast cancer. Firstly, the researchers identified 1,000 women located within the town and the surrounding area and developed a new breast cancer case for six years. Secondly, the researchers find another 1150 women located in the same region without breast cancer about the same age. Thirdly, participants from both group are asked about their personal, lifestyle, and reproductive pasts, including information about factors known to be associated with breast cancer (e.g., body mass index, drinking, smoking, menstrual and menopause history, use of hormone replacement therapy, birth control, family history). Besides, they are also asked about the recent jobs and past jobs. This information is allowing researchers to figure out likely exposures to cause breast cancer.
It usually uses for long incubation periods and rare diseases when conducting the case-control study. However, it also has limitations regarding the case-control study. For example, there is a probability that the conclusion drawn from the result is wrong as participants may recall partial or wrong memories in the past.
In conclusion, we discuss three observational studies: the cross-sectional study, cohort study, and case-control study. They all have different limitations that may lead to false conclusions. Therefore, when conducting such an experiment, we need to minimize the probability of certain biases.
Reference:
- Case control study. (n.d.). Retrieved April 06, 2021, from https://www.iwh.on.ca/what-researchers-mean-by/case-control-study
- Cohort study. (n.d.). Retrieved April 06, 2021, from https://himmelfarb.gwu.edu/tutorials/studydesign101/cohorts.cfm
- What is a cross-sectional study? — definition with examples. (2021, February 08). Retrieved April 06, 2021, from https://www.questionpro.com/blog/cross-sectional-study/#:~:text=Another%20example%20of%20a%20cross,geographical%20locations%2C%20and%20social%20backgrounds.
- Statistical studies. (2020, February 28). Retrieved April 06, 2021, from https://statisticsguruonline.com/statistical-studies/
