Email logo | Skype logoProject.cognibrain

   mobile logo+91-9940084145 | facebooktwitterlinkedingoogle+

Statistical Errors to Avoid In your Medical Thesis Writing

4 Statistical Errors to Avoid in your Medical Thesis Writing

Statistics is a tool that helps you assess the relationship between two or more variables and evaluate the study questions. Bio-statistics utilises this tool to resolve problems in biomedical sciences. It enables the researchers to analyse the effectiveness of a new drug, know the causal factors of a disease, calculate the life expectancy of the patient, the mortality and morbidity rate in a population, etc.

Though statistics is a primary tool in biomedical research, its misuse and abuse are common. And this misuse of statistics is one of the main reasons your medical thesis gets rejected.

So, here we explain 4 common statistical errors which you must avoid in your medical thesis writing:

1) Lack of clarity in presenting the statistical data

A cross-sectional study involving the students and faculty of medical colleges was carried out. The study results are:

  • 87% called statistics to be very difficult
  • 9% could not define P value
  • 45% could not define the standard deviation
  • 97% failed to calculate the sample size

Thus, it’s important to have a basic knowledge of statistics before writing a medical thesis. You must analyse and present your data clearly in your medical thesis.

2) Exceptional emphasis on data rather than the theory

The basic research studies carried out in colleges involve biochemistry, behavioral science, animal models, and cell cultures. This makes their statistical analysis challenging. Often the students statistically analyse their data only after performing their experiment; a strategy which provides only limited insight into the research study.

3) Poor decision-making before data collection

Following from the above point, statistical analysis should be planned before the data collection, like when deciding the sample size. This is so because the sample size can affect several variables in the outcome of the study.

Thus, it’s recommended to perform sample size computations for each outcome and then decide on the largest practical sample size. By doing so, you can avoid false-positive relationships in your results.

4) Biases in data collection and statistical analysis

While designing your studies, you must pay attention to the control groups (conditions), randomization, blinding, and replication. If you use a large sample size, randomization will help you avoid unintentional bias and confounding errors. Otherwise, it’s easy to confound and bias your study results.

How will statistical errors affect your medical thesis?

Just like the correct handling of data is important to get accurate results, statistical soundness is indispensable to avoid thesis rejection. The common statistical errors that occur during medical thesis writing can be classified into 4 categories:

  • Errors in study design (For example, no randomization in controlled trials; inappropriate control group)
  • Errors in data analysis (For example, incorrect reporting P values)
  • Errors in data presentation (For example, improper defining of standard deviation; pie charts to present distribution of continuous variables; no adjustment for multiple comparisons)
  • Errors in data interpretation

Errors in data presentation are easy to revise. However, if there are errors in data analysis, interpretation, or discussion of results, you need to make extensive changes throughout your thesis. In contrast, errors in study design can’t be rectified without repeating the whole study.

 How to avoid misuse of statistics in medical thesis writing?

Since statistical data is vital to the advancements in biomedicine, there have to be conscious efforts to avoid misuse and abuse in gathering, analysing, and presenting the statistical data.

Therefore, students must make themselves aware of ICMJE and ‘Statistical Analysis and Methods in the Published Literature (SAMPL)’ guidelines before beginning their studies. And also educate themselves about the best practices related to statistical methods.

Leave a Reply

Your email address will not be published. Required fields are marked *