I have decided to choose 10 words from the first 20 pages of each newspaper at random on one day. I have chosen 10 words from each page from the first 20 pages. 20 multiplied by 10 gives 200 words so this is what I would want to collect to analyse later. I will do this over a 5 day period so I reach my target of achieving 1000 for each newspaper. I could choose 10 words from 50 pages in the newspaper but some newspapers don’t tend to have more than 20 pages on a particular day and it might cause problems so I thought it would be best if I just chose the first 20 pages to sample. Also, I think that the first 20 pages would be an overview of the whole newspaper because the front pages are quite similar to the back pages and it wouldn’t be that different because tabloid and broadsheet newspapers tend to make the front as eye-catching as the front. The front tends to have real life story information and the back tends to have sports on, both would tend to have similar sized text as observed by me as I read the newspaper quite a lot. Near to the end, a typical newspaper would have information of houses so it would be quite hard to collect word samples from these so I would not like to do that so I am choosing to sample the first 20 pages. I will choose the words at random, this means that I would go onto the page that I am going to choose 10 words from and close my eyes. I will point to any word with the tip of my fingernail and the word that is under there will be chosen and tallied in my table in the correct group. If I place my finger in the middle of two words, I will do it again and choose another one until I land on just one word. I will do this until I get a total of 10 words on the page then move onto the next page. I will do this with each newspaper over a 5-day period Mon-Fri so I analyse 1000 from each newspaper. This kind of sampling is called “Simple Random sampling.” This way is not perfectly accurate but still I can use as my investigation is not a large scale one where this would affect the results. Some people would say that this way, your tend to choose sentences in the middle of the page. But in my opinion I don’t think this is true but I will bare that in mind.
I have decided to collect 200 words from Monday to Friday so it gives me a total of 1000 words. In order to achieve the total number of words from each newspaper I will have to collect 1000 words. I could’ve collected all 1000 words in one day but it wouldn’t represent the newspaper fairly. This is because, if on a day, the newspaper is using very long words and they usually use fairly small words then the results I have collected were biased. It wouldn’t be accurate, as it hasn’t treated the newspaper fairly. By fair I mean, the collected data was false and inaccurate so a variety of days needs to be chosen. Also, the newspaper might have chosen to use fairly small words on a day which could cause problems. To solve this problem, I will do this over Monday to Friday so any anomalous results don’t get noticed that much and the results would still be reliable. By reliable, I mean the data could be trusted.
Planning of hypothesis 2
“The Metro” has fewer words in a sentence than in “The Daily Mail”
I will sample 400 sentences in total from each newspaper.
I have chosen 400 sentences because I would want to choose a number of sentences, which I could draw a decent conclusion with. I don’t wan to only take a sample of 40 sentences, as they could be misleading. By accident, I could choose 40 long sentences or 40 short sentences. If that happened then the investigation would be pointless. This would most probably be just an overall summary of the page rather than the whole newspaper. I would want a decent number of sentences to investigate which I could use to figure out a decent conclusion to prove of disprove the hypothesis. If I chose 4000 sentences or some ridiculous like that in which I examined nearly every sentence in the newspaper then this would make the results more accurate but it would take so much time then I wont have time to analyse them.
I will group these results.
The results I will obtain could vary from one word to possibly 100. So I will have to group these because I wouldn’t like a table 2 pages in length with 100 groups from 1 to 100. The data I am going to collect is given numerical values and is called quantitative data. The number of words in a sentence is a discrete quantitative data. I will decide on which groups I will use to collect the data after I start the survey. This is because; I still don’t know the maximum and minimum number of words there will be in a sentence, in each newspaper. With these results I could plot pie charts or bar charts. I could also possible plot a histogram. The bar chart would show the size of each category/ group. With the pie chart, it would be a good way to compare the size of different groups. But with the histogram, it is like a bar chart but when you have unequal groups in the first place, I can’t just put then in so I’ll have to draw a histogram. I could also work out the mean, mode and median as an extra statistics tool.
I have decided to collect 4 sentences from the first 20 pages of each newspaper at random on one day. I have chosen 4 sentences from each page from the first 20 pages. 20 pages multiplied by 4 sentences equals 80 sentences. This is what I would like to collect one a day per each newspaper. I could choose 10 sentences from 50 pages per newspaper, but I would like to do this over a 5-day period rather than collecting all of the results on one day. I am choosing the first 20 pages because some newspapers don’t tend to have more than 20 pages on some or a certain day and it might cause problems so I thought it would be best if I just chose the first 20 pages to analyse. This is because, The Metro and the Daily Mail has a minimum of 20 pages every day so it won’t be a problem. Also, I think that the first 20 pages would be an overview of the whole newspaper because the front pages are quite similar to the back pages and it wouldn’t be that different because tabloid and broadsheet newspapers tend to make the front as eye-catching as the front. The front tends to have real life story information and the back tends to have sports on, both would tend to have similar sized sentences as observed by me as I read the newspaper quite a lot. Near to the end, a typical newspaper would have information of houses so it would be quite hard to collect word samples from these so I would not like to do that so I am choosing to sample the first 20 pages. I will choose the words at random, this means that I would go onto the page that I am going to choose 4 sentences from and close my eyes. I will point to any word with the tip of my fingernail and I will move to the start of that sentence. After that, I will count the number of words there are in that sentence including every word (no exemptions). For example, if the number of words in that sentence is 14, I will place a tally in the group would hold the data of 14. the group could be something like 10<sentence<20. I will tally the result. If I place my finger in the middle of two sentences or in a blank space, I will do it again and choose another one until I land on just one sentence. I will do this until I get a total of 4 sentences on the page then move onto the next page. I will do this with each newspaper over a 5-day period Mon-Fri so I analyse 400 sentences from each newspaper. This kind of sampling is called “Simple Random sampling.” This way is not perfectly accurate but still I can use, as my investigation is not a large scale one where this would affect the results. Some people would say that this way, you tend to choose sentences in the middle of the page. But in my opinion I don’t think this is true but I will bare that in mind.
I have decided 80 sentences from on each day from Monday to Friday so it gives me a total of 400 sentences. In order to achieve the total number of sentences from each newspaper I will have to collect400 sentences and this would have to represent the newspaper on any day. I could’ve collected all the results on one day but this wouldn’t represent the newspaper fairly. This is because, if on a day, the newspaper is using very long sentences and they usually use very small sentences then the results I have collected are biased. It wouldn’t be accurate, as it hasn’t treated the newspaper fairly. By fair I mean, the collected data was false and inaccurate so a variety of days needs to be chosen. Also, the newspaper might have chosen to use fairly small sentences on a day, which could cause problems. To solve this problem, I will do this over Monday to Friday so any anomalous results don’t get noticed that much and the results would still be reliable. By reliable, I mean the data could be trusted.
Planning of hypothesis 3
“The Metro” has bigger pictures (area) than in “The Daily Mail”
I will collect 100 pictures altogether from each newspaper (areas)
I think that an absolute minimum of 30 pictures to draw some kind of a conclusion. I have chosen 100 pictures because I would want to choose a number of pictures that I could draw a decent conclusion with. I don’t want to only take a sample of 40 pictures, as they could be misleading. By accident, I could choose 4 small pictures or 4 big pictures. If that happened then the investigation would be pointless. This would most probably be just an overall summary of a page rather than the whole newspaper. I would want a decent number of sentences to investigate which I could use to figure out a decent conclusion to prove of disprove the hypothesis. If I chose 1000 pictures or some ridiculous like that in which I examined nearly every sentence in the newspaper then this would make the results more accurate but it would take so much time then I wont have time to analyse them. Also, I would have to analyse over 30 papers just to look at that amount of pictures. I would then have to work out the areas than tally them onto a table, which would take a ridiculous amount of time.
I will group these.
The areas I will obtain could vary from 4cm3 to 400cm3. So I will have to group these because I wouldn’t like table 5 or 6 pages in length with 100 groups from 1cm3 to 400cm3. The data I am going to collect is given numerical values and is called quantitative data. The area of a picture is continuous quantitative data. I will decide on which groups I will use to collect the data after I start the survey. This is because; I still don’t know the maximum and minimum areas there will be of any picture, in each newspaper. With these results I could plot pie charts or bar charts. I could also work out the mean, mode and median of the areas. I could also plot a histogram. The bar chart would show the size of each category/ group. With the pie chart, it would be a good way to compare the size of different groups. But with the histogram, it is like a bar chart but when you have unequal groups in the first place, I can’t just put then in so I’ll have to draw a histogram.
I will analyse the first 20 pictures in each newspaper on a day.
I have to analyse the first 20 pictures on a day in each newspaper. This kind of sampling is systematic. 20 pictures multiplied by the 5 days equals 100 pictures. This is what I would like to collect one a day per each newspaper. I could choose 50 pictures a day per newspaper, but I would like to do this over a 5-day period rather than collecting all of the results on one day. I am choosing the first 20 pictures because some newspapers don’t tend to have more than 20 pictures on some or a certain day and it might cause problems so I thought it would be best if I just chose the first 20 pictures to analyse. This is because, The Metro and the Daily Mail definitely has a minimum of 20 pictures every day so it won’t be a problem. Also, I think that the first 20 pictures would be an overview of the whole newspaper because the front pages are quite similar to the back pages and it wouldn’t be that different because tabloid and broadsheet newspapers tend to make the front as eye-catching as the front. The front tends to have big pictures and the back tends to have similar large pictures on, both would tend to have similar sized pictures as observed by me as I look at the pictures in newspapers quite a lot. Carrying on from the first pages, the pictures tend to get smaller throughout to the end so I think that the first 20 pictures would represent the whole paper on that day. I will work out the area of the picture by multiplying the length and width of a picture. Then I would add a tally to the group, which contains that area. For example, if the length of the picture was 5cm and the width was 7cm then 7x5=35cm3. This would go into a category like 30cm3<picture size<40cm3.
I will do this over Monday to Friday so the total pictures are 100 pictures.
In order to achieve the total number of pictures from each newspaper I will have to collect 100 pictures and this would have to represent the newspaper on any day. I could’ve collected all the results on one day but this wouldn’t represent the newspaper fairly. This is because, if on a day, the newspaper is using very big pictures like they did on September 12th a few years ago when the twin towers fell down. But the newspaper usually uses very small pictures then the results I have collected are biased. It wouldn’t be accurate, as it hasn’t treated the newspaper fairly. By fair I mean, the collected data was false and inaccurate so a variety of days needs to be chosen. Also, the newspaper might have chosen to use fairly small pictures on a day, on which or the day before nothing exciting didn’t happen. To solve this problem, I will do this over Monday to Friday so any anomalous results don’t get noticed that much and the results would still be reliable. By reliable, I mean the data could be trusted.