25-03-2022
# Dansk litteratur set fra skyen, 📚☁️📊

## Three main findings in the Bog.nu-dataset

A rich dataset with book reviews published in Danish media in the years 2010-2021.

| 1. gender bias | 2. grade inflation |
:-------------------------:|:-------------------------:
 |
**3a. divided readers I** | **3b. divided readers II**
 | 

## Gender findings

#### Gender of reviewer x Gender of author:
Female reviewers are grading female authors the highest.
Male reviewers are grading female authors the lowest.

#### Review-gender is not equally distributed across media types:

Male reviewers are dominating in newspapers.
Female reviewers are dominating in blogs.



#### And there is a difference in grading across mediatypes:
The highest grades are given in Blogs and in 'Online media without quote'.

The lowest grades are given in newspapers and regional newspapers.


#### How does female and male reviewers grade in different media types?
Female reviwers are giving low grades in newspapers. Actually the lowest grades are given by female reviewers in newspapers.


#### Question: Is the found gender bias due to difference in grading cultures and distribution of review-gender across different media types?

## Grade inflation

#### However, as shown above, there is a difference in grading in different media types.

#### From ~2015 we see a rapid increase in number of blog reviews



## Divided Readers
### book clusters

Female and male reviewers are reading (or at least reviewing) very different books.

Most books are reviewed less than 10 times.

Most books are either read by female reviewers **or** by male reviewers.


<center><img src="https://i.imgur.com/7aHKeEy.png"></center>