Now I will use these results in the forecast formula.
R = 25 – ( 110 / 10 )
R = 25 – 11
R = 14
As you can see, this is not between 7-10, far from it in fact. This shows me that the forecast formula is very inaccurate, and therefore unreliable. I will test the forecast formula against another book, but now I must use the syllables formula. To use this formula, I must first know the mean number of words per sentence (known as W) and the total number of syllables in 100 words (known as S). I found the value of S in this case by multiplying the frequency of syllables per word, by the amount of syllables per word that they had.
That would look like this:
1 x 73 = 73
2 x 24 = 48
3 x 6 = 18
4 x 0 = 0, the sum of this is 139, and therefore the value of S.
I will find the value of W by adding the number of words in the sentences, and then dividing this sum by the number of sentences. The calculation is as follows:
19
10
10
5
5
25
14
5
13
16
6
3
7
+8
From this I can tell that the syllables formula is more reliable than the forecast formula, as the results of the syllables formula fall between the reading age ranges the publisher has set out.
In order to check whether these formulae have been reflected well in this test, I have decided to test them against another book, of which the publishers have given a range of reading ages. I will use the book ‘Harry Potter and The Philosophers Stone’, simply for the fact that I know what the publishers have stated the reading age range as, in this case it is 9-12.
Here are my results:
I will now see what forecast formula shows the reading age of this Harry Potter book to be:
W= 76 / 4 = 19
S=( 1 x 66 ) + ( 2 x 30 ) + ( 3 x 3 ) + ( 4 x 1 ) = 139
R = 25 – ( 99 / 10 ) = 25 – 9.9 = 15.1
I can now tell that the forecast formula is flawed, because the reading age it comes up with is far outside the range 9-12. Now I will see what the syllables formula shows:
R = 2.7971 + ( 0.0778 x 19 ) + ( 0.0455 x 139 ) =10.59
From this I can see that the syllables formula is the most accurate, and therefore I will use this to find the reading age of the broadsheet and tabloid newspapers.
However, the syllables formula doesn’t include the amount of syllables per word as criteria, so I have made my own formula which incorporates this. This is how I worked my own formula out.
First, I have to determine what criteria I am using. I know that because the syllables formula does not factor in syllables per word, I will use that instead in my formula. So far, my formula looks like this:
R = (total number of syllables in 100 words / 100)
However, it is far from complete. This would give very strange results, so I must raise the number I get given as the reading age. To find what number to multiply it by, I should use a reading age that I have already worked out to help me work out what the multiplier should be. This is how I will do it:
R = 10.59
10.59 = (139 / 100) = 1.39
10.59 / 1.39 = 7.61 (the reason I divided the reading age by the total number of syllables divided by 100 (abbreviated as TNSDB100) is because I needed to know what to multiple the TNSDB100 by to get the reading age.)
Therefore my formula could be R = (total number of syllables in 100 words / 100) x 7.61. I will test this again to see if the results are reasonable. (I have used data from the Times, below)
R = 143 / 10 x 7.61 = 1.43 x 7.61 = 10.88
This is a reasonable formula, as this is very close to what the syllables formula came up with, and as previous results have shown, the syllables formula is also reasonable. I believe that it is suitable for me to use in my analysis, and I will refer to it as the reasonable formula.
Now that I have two formulae to work with, I will start to compare the reading ages of broadsheet newspapers, and tabloid newspapers. The following table shows the amount of syllables for each word:
The Times is the broadsheet, and the Daily Mail is the broadsheet.
This time, I have only included ‘out of 100’ and not ‘out of 150’, because ‘out of 150’ is now irrelevant because I am no longer using the forecast formula. I will work out the reading age for the broadsheet newspaper (the Times) using the syllables formula.
R = 2.7971 + ( 0.0778 x 28.6 ) + ( 0.0455 x 143 ) = 11.52
I will now see what the reasonable formula comes up with.
R = 143 / 100 x 7.61 = 1.42 x 7.61 = 10.8
I will now see what the syllables formula and the reasonable formula find the reading age of the tabloid newspaper to be, respectively.
R = 2.7971 + ( 0.0778 x 20.1 ) + ( 0.0455 x 132 ) = 10.36
R = 132 / 100 x 7.61 = 1.32 x 7.61 = 10.04
From this, I have deduced that my hypothesis is correct, because, according to these formulae, the reading age of a broadsheet newspaper is higher than the reading age for a tabloid newspaper.
To provide further evidence, I will use averages to show differences between the newspaper types. I will first show the how many words per sentence that there is for both.
For broadsheet: 27 + 37 + 35 + 25 = 124
124 / 4 = 31
For tabloid: 25 + 19 + 15 + 15 + 22 + 14 + 30 = 140
140 / 7 = 20
This shows me that broadsheets have a higher amount of words per sentence, which could mean that tabloid readers need articles requiring a shorter attention span, as generally, shorter sentences keep one’s attention better, and/or it could suggest that broadsheet readers have higher attention spans.
Now I will show the amount of syllables per word:
For broadsheet: 142 / 100 = 1.42 syllables per word
For tabloid : 132 / 100 = 1.32 syllables per word
This shows me that, generally there are more words with more syllables in the broadsheet newspaper. The amount of syllables in each word can be a general indicator of how large the word is, and how large a word is often suggests how much harder it is to comprehend.
Conclusion
What my results showed me: from my results, I could see that a broadsheet newspaper required a higher reading age than a tabloid newspaper, although by a very small margin, with the reasonable formula showing a difference of 0.76 and the syllables formula showing a difference of 1.16. This amount is negligible, and so it showed me that the gap between the reading ages of broadsheet and tabloid is in fact quite a lot smaller than I had previously thought, although, if I had selected a different passage for either of them, my results could have been very different. My mean amount of words per sentence could suggest that the tabloid reader has a shorter attention span than a broadsheet reader, which does not necessarily indicate that tabloid readers are of a lower intelligence than broadsheet readers. Also, there are more syllables per word in the broadsheet newspaper. The amount of syllables in each word can be a general indicator of how large the word is, and how large a word is often suggests how much harder it is to comprehend, which does suggest that the people who read the broadsheet are more intelligent, or that the people who read tabloids prefer to have less big words in their newspaper.
Sources of error: in any experiment, there are errors, and mine is no exception. My formula does not take into account many of the other criteria, such as words per sentence, word length, % of words with, for example, more than 3 syllables. It is also hard to judge how important and relevant some of the criteria is, so maybe I chose the wrong one to focus on. About the passages I chose, and how I took data from them: the passages I chose did not end with a full stop, they both stopped in the middle of a sentence, and so when I took the mean amount of words per sentence, it may have been the slightest bit inaccurate as I could not include the full 100 words. When I was counting the words, I counted most hyphenated words as one, such as counting ice-cream as one word, whilst I know some colleagues counted ice-cream as two words. This could make comparison a slight bit skewed, emphasis on slight, and such cases were not common. I may have incorrectly counted the words, and maybe my pronunciation of certain words could have meant one word got less or more syllables than it was supposed to, and this could make comparison to others inaccurate. I also included proper nouns, which may have been unfair, as my broadsheet passage could have included very long names, as could the tabloid passage, which may have affected the average number of syllables per word and the reading age. There was also an element of human error, such as miscounting words, making errors in calculations, etc.
Improvements: I could have used data from a whole newspaper, I could have used a computer or something which could have helped me keep count of words, I could have spent more time developing a formula, incorporating many forms of criteria, such as cumulative frequency of syllables per word, etc, I could have selected a passage which didn’t end in the middle of a sentence. I could have compared the front pages of the newspapers, to see if there was any difference there, or lack of one, and I could have compared the length of articles.
