Statistics Project

Statistics Project

I aim to compare mass-appeal tabloid newspapers and quality newspapers

by attempting to find statistical differences. To represent the

mass-appeal papers, I chose the Daily Mirror and for the text-quality

based newspapers, I chose the Times. Hopefully, there will be some

significant statistical differences in the style of journalism which I

will be able to comment on.

Pre-Test

Data Collection: I decided to choose similar pages from both the Times

and the Mirror with roughly equal numbers of paragraphs and adverts,

pages 4-5, or 4-6, as in the Mirror there were not enough sentences to

take samples from. To find mean sentence lengths in the two papers, I

decided I would sample systematically from my populations, counting

the number of words in every 3rd sentence. I came up on several

problems quickly - should I include headlines in my count? I decided

against it, as headlines tend to be shorter than normal sentences. The

next problem came with numbers - did they get counted as words in the

sentences? Making sure that I did the same with both papers, I decided

to exclude numbers in my count. I also decided to exclude any

sentences in adverts, as the number of adverts on the compared pages

varied. I then took a mean and found the standard deviation of my

data.

To find the average number of words per sentence, I decided to

'cluster-sample', and count the first 30 words in the first paragraph

of page 4 in each paper. I decided that I would again exclude numbers,

and that hyphenated words counted as a single word. Again, when I

found all the data, I found its mean and the standard deviation.

Location: As can be seen from this box and whisker diagram, the Times

has a similar mean sentence length (20 to 1sf.) to the Mirror (18).

This shows that the average sentences in the Mirror and the Times

contain a similar number of words. In the box and whisker diagram for

word lengths, it is visible that their medians are the same.