Lecture notes. Written from the material of GTPB Psls20, licensed CC BY 4.0. These are notes, not a transcript: the material has been reorganised and rewritten. This adaptation carries the same licence, and the original is linked above.
42. Statistics and Data Exploration Review¶
What this covers¶
This chapter is the first wrap-up session of the course, placed right after the material on experimental design and on the basic concepts of statistics. It introduces nothing new: it revisits the three strands of Part 1 — experimental design, data exploration, and statistical inference — and turns each one into a question the reader should be able to answer about their own understanding. It assumes the lectures on experimental design, on random variables and estimation, and on data exploration that came before it.
What Part 1 covered¶
The wrap-up organises Part 1 of the course into three strands:
- Experimental design
- Data exploration
- Statistical inference
Experimental design¶
The concepts singled out for review:
- Defining the population under study
- The importance of randomisation
- Sample size
- Control: simultaneous control groups, randomising, and blinding
- Confounding — and recognising which kinds of studies carry the higher risk of it
Basic concepts of statistics¶
- Types of variables
- Random variables
- Probability theory
- Statistical modelling
- Estimation
- Why every statistic or estimator computed from data is itself a random variable
- The distinction between an empirical approach and one that rests on distributional assumptions
Data exploration¶
The wrap-up names data exploration as the third strand of Part 1 but does not add further material of its own under this heading — the closing questions on showing data raw and on spotting departures from normality are what stand in for it here.
Questions to check your own understanding¶
The wrap-up closes with five questions, used to gauge confidence topic by topic rather than to grade it. The point is to answer them honestly to yourself before moving on, not to score them:
- How confident are you in recognising different types of variables in a study?
- How confident are you in understanding the importance of randomisation?
- How confident are you in understanding how data, statistics and conclusions change from sample to sample, and what the impact of sample size is on that uncertainty?
- How confident are you in explaining to somebody else the importance of data exploration, and the importance of showing the data as raw as possible?
- How confident are you in spotting data that deviates from the normal distribution?
Sources¶
docs/omics-statistics/gtpb/psls20/theory/wrapup/wrapup1.md— converted fromtheory/wrapup/wrapup1.Rmd(GTPB PSLS20 course, CC BY 4.0). This is the only input supplied for this chapter; no slides, transcript, or exercise set accompanied it.- The wrap-up itself points back to the earlier lectures on experimental design, basic statistical concepts, and data exploration; that material is not contained in this file and is not reproduced here.