# PhD Course Basic Statistics for health researchers

## Table of Contents

## Welcome

This is the course homepage of the English version of the PhD course
Basic statistics for health researchers. The course description,
including the **learning objectives**, is available from the website of
the PhD school, here:

The webpage will be **updated regularly** during the course. Check the
footnote at the bottom of the page for the date and time of the last
update.

## Place & Schedule

The course takes place at the Center for Sundhed og Samfund - CSS - Ă˜ster Farimagsgade 5, 1014 Copenhagen. Classes run from 8:00 to 15:00. The room numbers and schedule are given in the following table:

Date | Day | Room (8:00-15:00) | Topics | Teachers |
---|---|---|---|---|

17 April 2023 | Monday | CSS-7.0.40 | Overview, data, descriptive statistics, concept of statistical inference, confidence intervals | Paul Blanche, Carolin Herrmann |

19 April 2023 | Wednesday | CSS-7.0.40 | Hypothesis testing, tests for continuous outcomes, multiple testing | Paul Blanche, Carolin Herrmann |

24 April 2023 | Monday | CSS-7.0.40 | Univariate linear regression, correlation, regression to the mean | Paul Blanche, Zehao Su |

26 April 2023 | Wednesday | CSS-7.0.40 | Analysis of Variance (One-way and Two-way ANOVA) | Paul Blanche, Zehao Su |

3 May 2023 | Wednesday | CSS-7.0.40 | 2x2 tables, odds ratio, two sample tests for binary responses | Paul Blanche, Zehao Su |

8 May 2023 | Monday | CSS-7.0.40 | Logistic regression | Paul Blanche, Zehao Su |

10 May 2023 | Wednesday | CSS-7.0.40 | Multiple linear regression, confounding, interaction | Paul Blanche, Alessandra Meddis |

15 May 2023 | Monday | CSS-7.0.40 | Repeated measurements | Brice Ozenne, Alessandra Meddis |

17 May 2023 | Wednesday | CSS-7.0.40 | Survival analysis | Paul Blanche, Alessandra Meddis |

24 May 2023 | Wednesday | CSS-7.0.40, CSS-7.0.06 | Presentation and discussion of homework assignments | Paul Blanche, Brice Ozenne |

## Lectures and R-demo

Lecture notes and accompanying R-demos should be available via the
links in the table below no later than **two days** before the course day.
Some R-demos use external data not already available in R
packages. Those are also provided via the table below.

Course Day | Handout (1x1) | Handout (2x2) | R-demo | External Data |
---|---|---|---|---|

1 | Lecture-1 | Lecture-1-2x2 | Rdemo-1 | none |

2 | Lecture-2 | Lecture-2-2x2 | Rdemo-2 | none |

3 | Lecture-3 | Lecture-3-2x2 | Rdemo-3 | th , ckd |

4 | Lecture-4 | Lecture-4-2x2 | Rdemo-4 | none |

5 | Lecture-5 | Lecture-5-2x2 | Rdemo-5 | dalteparin, SCD |

6 | Lecture-6 | Lecture-6-2x2 | Rdemo-6 | Framingham |

7 | Lecture-7 | Lecture-7-2x2 | Rdemo-7 | VitaminD |

8 | Lecture-8 | Lecture-8-2x2 | Rdemo-8 | none |

9 | Lecture-9 | Lecture-9-2x2 | Rdemo-9 | subfertile, carcinoma, HFactionHosp |

The course is not based on a single **textbook**, but the following book
is an excellent reference. It covers most of the topics of the
course (often in more details) and has been written for a similar
audience to that of the course. *Regression with linear predictors*,
by Per Kragh Andersen and Lene Theil Skovgaard (Springer, 2010).

## Additional reading and video material

To complement the lectures and practicals, for each course day we recommend short videos and papers, to watch/read preferably before or after each course day. This material is independent of the lectures and practicals. It has been developed by other teachers than those of the course.

We recommend this material because we find it both **entertaining**
and **educative**.

When you watch/read before the course day as recommended, it should help you to better follow the lecture. When you watch/read after the course day, it should help you to revisit some important concepts and/or further learn on a few selected topics.

## Practicals

Practical experience with data analysis and statistical methods will be learned with an emphasis on understanding of the output of the statistical software and the interpretation of the results. Solutions for the exercises are provided only for the statistical software R.

Course Day | Exercise | External Data | R code solution | Full Solution (R code & answers) |
---|---|---|---|---|

1 | Exercise-1 | SCD | Solution1.R | Solution1.pdf |

2 | Exercise-2 | none | Solution2.R | Solution2.pdf |

3 | Exercise-3 | SCD | Solution3.R | Solution3.pdf |

4 | Exercise-4 | SCD | Solution4.R | Solution4.pdf |

5 | Exercise-5 | smoking , sedative | Solution5.R | Solution5.pdf |

6 | Exercise-6 | MI | Solution6.R | Solution6.pdf |

7 | Exercise-7 | BW , Brain | Solution7.R | Solution7.pdf |

8 | Exercise-8 | none | Solution8.R | Solution8.pdf |

9 | Exercise-9 | colon2 | Solution9.R | Solution9.pdf |

## How to prepare for the course

### Internet

You will work with your own laptop. You will need Internet during the course. It is your own responsibility to be able to connect.

### Statistical software

The focus of this course is not on how to use statistical software. But statistical software is needed for all data analyses and examples that illustrate the statistical methods. The free statistical software R will be used throughout the course, via R-studio (https://posit.co/downloads/). It is expected that the students learn the syntax and semantics of R before and during the course by themselves. Note that this will often mean a lot of extra hours for preparation and self-training in addition to the actual teaching hours.

**All students are expected to start working with R syntax and semantics**
**several weeks before the course starts**. **A minimum level** corresponding
to that obtained after completed our online introduction to R at
https://biostat.ku.dk/r/ is considered as a **prerequisite**.

In this introduction, we guide you through how to install R, how to load data, data manipulation and simple calculations and plots. Estimated number of hours to complete the introduction: 10-15 hours depending on your R- and technical skills.

The participants are expected to use their own laptops during the course, to have installed all relevant software and to have downloaded all data for use during the course.

### R packages

To run the code of the R-demos as well as to solve the exercises with
R, specific packages need to be installed. Below is the list of
packages needed for each course day, **in addition** of those for the
previous days. **Please install the packages before the start of each
course day**. The list will be updated no later than two days before
the course day.

Some of the packages are essential to use specific functions, others are used to load example data.

Course Day | R packages |
---|---|

1 | DoseFinding , MESS , timereg , HistData |

2 | nlme , coin |

3 | none |

4 | multcomp , sandwich, HSAUR2 |

5 | Publish |

6 | none |

7 | none |

8 | reshape2, LMMstar, ggplot2, mice , nlmeU |

9 | survival , prodlim, survRM2 |

## How to pass the course

To pass the course you have to

- attend 80% of all teaching units (we count the signatures).
- turn in your homework assignment in due time.
- present the results of your homework on the last course day.

## Homework assignment

A homework assignment is handed out after lecture 5. Participants work with their own data or data related to their own research provided by their PhD supervisor. The homework assignment is turned in after lecture 9.

The homework assignment is now available **here**.