As part of my first year internship, investigating emotional processing in OCD with comorbid depression using task-based fMRI, I conducted several statistical analyses for my research project. The following R scripts are included:
- Correlation_madrs_roi.R - Correlational analysis between MADRS scores and different ROIs in OCD subjects
- YBOCS_MADRS_normality.R - Test for normality of YBOCS and MADRS scores
- Clinical_variable_analyses.R - Descriptive statistics of clinical variables (age, gender, MADRS scores, OCD severity) in OCD subjects and healthy controls
- Correlation_madrs_ybocs.R - Correlational analysis between MADRS and OCD severity
As part of my 'Statistics in Neuroscience' course in the first year of my masters, we were required to submit a data portfolio demonstrating our implementation of statistics in R based on the topics we covered. I was provided with a dataset which included a set of (simulated) variables pertaining to factors that influence the risk of develping diabetes. This dataset was used to formulate a research question, which I was able to answer by carrying out various types of statistical analyses explored in this course.
The final portfolio consisted of two components:
- A poster, as suitable for presentation at a scientific conference
- The R script used to explore my dataset and generate results shown in the poster
FINAL GRADE: 8.82/10
Link to poster: https://drive.google.com/file/d/1GdqW8GbEbOXNsVYTMA88a6EKmqWBHf6p/view?usp=sharing
Data portfolio instructions:
A group of researchers is interested in what factors influence risk of developing diabetes. They recruited 150 adults from the general population using newspaper and online advertisements. The participants were screened to exclude patients with pre-existing diabetes. The researchers measured participants’ blood glucose levels (mg/dL), which are an indicator of risk for diabetes. Measurements were taken 3 times over the course of one day: a fasting glucose measure immediately after waking (when levels are the lowest), a measurement after the midday meal, and a measurement just before bedtime. Participants were also given a self-report survey in which they answered a number of questions on demographic characteristics, diet, and exercise.
The researchers’ aim is to identify what factors put people at risk for developing diabetes, as indexed by higher blood glucose levels. They also want to examine how these risk factors are associated with patterns of change in glucose levels throughout the day.
What conclusions can you draw about diabetes risk factors and blood glucose levels based on this dataset?