I divided the number of words I still required by two giving me two counts - if this sum gave me a fraction I rounded it up for the first count and down for the second (e.g. if I required thirteen words, I divided 13 by 2 giving 6.5 which I rounded to 7 for the first count and 6 for the second).
The first count represented the number of paragraphs at the start of the article from which I needed to take another word. I took the next word from each of those paragraphs.
The second count represented the number of paragraphs at the end of the article from which I needed to take another word. I took the next word from each of those paragraphs.
At that point I had the thirty words that I required from each article.
I avoided taking the first word from a paragraph as most sentences in the English language start with a small word such as ‘it, the or a.’ Paragraphs in both the newspapers would be equally likely to start this way and therefore would bias my results if included.
I excluded names and numbers from my selected words as these were likely to be the same in both articles. I also excluded hyphenated words as these tend to be easier to read.
If one of the words I selected was a name, number or hyphenated word, I selected the word prior to it in the paragraph until I found a word that was selectable.
Results
The following paragraphs show the results I obtained for each article pair.
For each article I have provided a tally chart of word lengths, a graph showing a comparison between the pair of articles and any observations I made. Also I calculated the averages.
First article
The article in the broadsheet was titled “US braced for storms as hurricane lashes Cuba” and contained twenty paragraphs.
The article in the tabloid was titles “Killer hurricane storming in” and contained 17 paragraphs.
Table 4.1 below is a tally chart for the above articles showing the frequency of each word length between 1 and 12.
Table 4.1 - Word length distribution
Table 4.2 shows the averages for both the broadsheet and tabloid.
Table 4.2 – Broadsheet and Tabloid Article Averages
As you can see from this table the averages are very similar.
* The mean for the tabloid shows a value of 3.9 but it is not possible to have a word length of 3.9. So I rounded it up to 4.
The graph shows a positive skew for both newspapers. This is not unexpected however as in the English language we tend to use shorter length words more often than the longer ones. At the moment it appears as though the broadsheet uses more of the shorter words than the tabloid, but this is only the first article and so you can’t tell for sure.
Second article
The article in the broadsheet was titled “Queen honours war’s heroines” and contained ten paragraphs.
The article in the tabloid was titles “Honoured at last the heroines of World War Two” and contained twenty-one paragraphs.
Table 4.3 below is a tally chart for the above articles showing the frequency of each word length between 1 and 12.
Table 4.3 - Word length distribution
Table 4.4 shows the averages for both the broadsheet and tabloid.
Table 4.4 – Broadsheet and Tabloid Article Averages
Again, the averages are very similar.
* These have been rounded to avoid having a value representing part of a word.
Again, the graph shows a positive skew for both articles, but in these articles the broadsheet used more longer words than in the previous article.
Third article
The article in the broadsheet was titled “Potter’s magic spell turns boys into bookworms” and contained thirteen paragraphs.
The article in the tabloid was titles “Potter coins it like Beckham” and contained thirteen paragraphs.
Table 4.5 below is a tally chart for the above articles showing the frequency of each word length between 1 and 12.
Table 4.5 - Word length distribution
Table 4.6 shows the averages for both the broadsheet and tabloid.
Table 4.6 – Broadsheet and Tabloid Article Averages
* These have been rounded to avoid having a value representing part of a word.
This graph shows a bimodal distribution in the broadsheet and the tabloid uses a larger range of word lengths than in the previous articles.
Fourth article
The article in the broadsheet was titled “Peitersen will have to wait for Thorpe’s finale” and contained fourteen paragraphs.
The article in the tabloid was titled “Pistol Piet vs Thorpedo” and contained nineteen paragraphs.
Table 4.7 below is a tally chart for the above articles showing the frequency of each word length between 1 and 12.
Table 4.7 - Word length distribution
Table 4.8 shows the averages for both the broadsheet and tabloid.
Table 4.8 – Broadsheet and Tabloid Article Averages
* These have been rounded to avoid having a value representing part of a word.
The broadsheet article is approaching a positive skew and the tabloid is almost a normal skew.
Summary
Each article on its own does not give a final result because it is only a small sample. To get a more accurate picture of what word lengths the articles use I decided to total the individual results and create the graph shown on the following page.
The overall word length comparison appears to show that the newspapers word length use is similar which was not my expected result.
As a final step, I totalled the use of words with length between 1 to 6 and 7 to 12 and produced a summary word length comparison graph. This showed that the broadsheet contained more of the shorter words than the tabloid and the tabloid contained more of the longer words.
This was completely opposite to my prediction. This could be for a variety of reasons including my selection of articles, my choice of newspapers, the size of the sample taken (i.e. only taking thirty words from articles that contained around six hundred words or more) and using Sunday papers that may be aimed at a different audience than the daily papers.
Table 4.9 shows the averages for the summary of the broadsheet and tabloid articles.
Table 4.9 – Summary Broadsheet and Tabloid Article Averages
* These have been rounded to avoid having a value representing part of a word.
Conclusion
I concluded that the broadsheet had the best readability. This is because it uses more shorter words than the tabloid. The tabloid uses more of the longer words as shown by the Summary Word Length graph.
This means that my hypothesis was incorrect because I predicted that the tabloid would use more of the shorter words.
I was quite surprised by theses results because the broadsheets are read more by professional people and the tabloids more by manual workers who may not have had such a good education.
