Planning
To prove my hypotheses I am going to select one newspaper article from the two papers. They will be of similar size to make the test as accurate and as fair as possible.
- The two newspapers I compare will be of the same date to make the investigation as fair as possible.
- The section I will compare on each paper will be on the same subject; this will give a balanced view on each paper and not bias towards one e.g. a medical topic using longer words compared to a sport topic, which would use shorter words.
- From each section I will use the first 100 words (not including the headline); this should be a large enough sample size to make the investigation fair.
- To show the word length on each section, I will count how many letters are in each word and record on a tally table for each newspaper, this will be the easiest and quickest way to record my results.
- I will plot the frequency results from each tally table on to a Frequency Polygon Graph so the results can be compared easily and clearly, because a tally table solely shows numbers
- To show the sentence length on each section, I will count and record how many words are in each sentence for each newspaper on a tally table and group my results, this again will be the easiest and quickest way to record the results clearly.
- I will show the frequency results from each tally table using a pie chart; this will show the information clearly as the total words per sentence will be quite large compared to letters per word and a pie chart will compare this better.
- From the information I have gathered and to prove my hypotheses I will calculate the Mean, Median, Mode and Range of:
Number of letters per word in both newspaper articles
Number of words per sentence in both newspaper articles.
Also I will plot my results on a Culminate Frequency Curve Graph to show the running total of the data collected and calculate the upper, lower and inter quartile range from this. I will then show these results on Box Plots, as this will be a good way of comparing the data. To use this method for the sentence length my data will be grouped so that it will be easier to use.
- I will analyse my results and compare them against both newspapers to see if my hypothesis was correct.
- Finally I will look at ways I could have improved my investigation.
Calculations
To show my results from the Tally Tables I used the following calculations:
-
Mean – total no. of letters/words ÷ total no. of words/sentences
-
Median – the middle number
-
Mode – the most common
-
Range – from the smallest to the largest.
The Sunday Telegraph
Number of letters per word
- Mean = 4.75
- Median = 4
- Mode = 3
- Range = 13
Number of words per sentence
- Mean = 26.25
- Median = 27
- Mode = Not Applicable
- Range = 36
News of the World
Number of letters per word
- Mean = 4.09
- Median = 3
- Mode = 3
- Range = 12
Number of words per sentence
- Mean = 19.5
-
Median = 10 → 15 group
-
Mode = 10 → 15 group
- Range = 25
To show my results from the Cumulative Frequency Polygon Graphs, I used the following calculations:
-
Upper quartile – the top 25% of the total data collected
-
Lower quartile – the lower 25% of the data collected
-
Interquartile range – the difference between the upper and lower quartiles.
-
Median – the middle value of the data collected
The Sunday Telegraph
Letters per word
- Upper quartile – 6.2
- Lower quartile – 2.2
- Interquartile range – 4
- Median – 3.8
Words per sentence
- Upper quartile – 35
- Lower quartile – 15
- Interquartile range – 20
- Median – 25
News of the World
Letters per word
- Upper quartile – 4.1
- Lower quartile – 2.2
- Interquartile range – 1.9
- Median – 2.9
Words per sentence
- Upper quartile – 28.8
- Lower quartile – 12
- Interquartile range – 16.8
- Median - 14