Module Information

Data handling and statistical analysis. This part of the course will provide students with an understanding of the different kinds of data generated by experimental science and of the most widely used statistical techniques. Emphasis will be placed on providing students with the knowledge base necessary to select the most appropriate statistical tests to differentiate between chance events and those more likely to be attributable to the factor being manipulated in the study. The range of tests covered will reflect the broad range of degree disciplines represented on the module, but will include, among others: T-tests (independent and paired), ANOVA (one-way and two-way), linear regression, various non-parametric tests, tests for normality, and selected multivariate analyses. For each test type, students will be given practical experience using mock data applied to different statistical software platforms.

Composing a tractable research plan. Students will be guided from the initial concept of an idea through to the assembly of a detailed and valid research plan. This process requires creative identification of a pertinent gap in scientific knowledge or a short-coming in methodological capability. Students will be guided on how to encapsulate their idea into the form of a tractable research question and then on how to convert this into a testable alternative hypothesis and associated null hypothesis. The importance of specifying both the dependent and independent variables during hypothesis formulation will be particularly emphasised. Students will be encouraged to consider how to measure the dependent variable and how to control (or group) the independent variable when scoping their research strategy. Help will be provided over the creative process of identifying possible constraints, sources of potential bias, assumptions and confounding variables/factors that could compromise the ability to test the hypothesis set. This then leads to specifying measures that mitigate for the effect of these factors when shaping the structure and scale of the experimental design. Consideration will be given to the number and genetic diversity of organism under study, the type of data generated (quantitative or qualitative), of the controls and (technical and biological) replicates, and of the statistical approach best suited to effectively differentiate between null and alternative hypotheses. Thus, an understanding of the various statistical analyses available is shown to be as important in the design of the experiment as it is in the interpretation of the results. A reiterative approach will be encouraged for the entire planning process. Hypothesis creation according to formal criteria will be encouraged to enable better differentiation between good and vague research topics. This will aide selection of dissertation topics for their final year and will foster a more sceptical perspective of research findings. The importance of selecting a dissertation topic that matches their career aspirations will be emphasised, along with the scope for exploitation of the results generated in a non-scientific context (public good/commercial/IP).