Assignment Question
Read All Thoroughly. The final assignment for this course is a project where you will complete an individual study including statistical analysis and visualization in R. The aim with this project is to let you practice and show your skills using different statistical methods and also to find, extract and handle data related to your specific field of interest (e.g. environmental, epidemiological or economical questions). Your task is to use your imagination to come up with four different research questions that you think can be answered with data available in the literature or in any electronically available form. As an alternative strategy you may also undertake an experiment/survey and compile data of your own. Then, you should perform the correct statistical analysis and do the interpretation of the results to answer your research questions. There are requirements that need to be filled for the report to be a pass, see details in Requirements below. One requirement is the chosen analyses, where you have to make sure that: At least one of the questions must be answered with a trend analysis, e.g. using correlation or linear regression. At least one of the questions must focus on differences between two groups, i.e. using t-tests or non-parametric alternatives in the analysis. At least one of the questions must focus on differences between more than two groups, i.e. using different versions of ANOVAs or non-parametric alternatives in the analysis. At least one of the questions must deal with count data, i.e using chi-square tests in the analysis.
More Instructions: Use the IMRAD-structure for the report. Make sure to use references when needed, using your preferred referencing system. This means having content in the parts Introduction, Method, Result, Discussion and References. Research questions should be presented in context with relevant background information and references. Research questions answered by statistical analyses using all the methods from the course, i.e. correlation/linear regression, t-tests (parametric and non-parametric), ANOVA and chi-square tests. Results described both in text and in visualizations. If data needs to be displayed in tables, create proper tables (preferably using R). Tabulated data in the form of direct output from R is not allowed. All figures and tables need to be numbered, and figures should have figure legends/captions (below the figures) and tables should have table heads/captions (above the tables). The legends and heads should concisely describe what the figure or table shows, and give the reader the necessary information to be able to interpret the figure/table. Furthermore, make sure to use proper, intuitive and well-formatted axis labels in your figures. Results should be discussed in a context with relevant background information and references. All R code is to be displayed in an appendix in the report, or in a separate supplementary file. Make sure that your code is easy to follow e.g. by separating the code into your different research questions. If you choose to hand in a supplementary file, make sure it is a the correct format (i.e do not just hand in the R Markdown-script). No code in the actual obligatory (IMRAD) parts of the report. Only keep relevant R code in the appendix! No irrelevant warnings or messages when e.g. loading packages should be displayed. If you have not before, download the R Markdown cheat sheet, from the Help-menu in RStudio. There you can learn how to format the output of different chunks using knitr syntax in the chunk header (e.g. {r warning=FALSE} turns off the display of warnings in the knitted report). The file format of the report should be either doc, docx, odt, html or PDF.
