Explore BrainMass

# hypothesis about the length of words in newspapers

Not what you're looking for? Search our solutions OR ask your own Custom question.

This content was COPIED from BrainMass.com - View the original, and get the already-completed solution here!

Context: It would normally follow on from work on setting up and testing hypotheses, and statistical analysis.

Question: Suresh is comparing magazines and newspapers.
He chooses a passage from one newspaper and one magazine. They each contain 100 words and he counts the lengths of all the words.

now this is what you have to do once you have read the text above.
1. Write a hypothesis about the length of words in newspapers and magazines,
2.Design and carry out an investigation to test your hypothesis.

When Doing this question make sure your look at these three things it is what needs you need to keep in mind when Answering the question, 1. Specify the problem and plan. 2.Collect Process and represent data. 3.interpret and discuss results.

You can do a scatter graph and draw a line of best fit or You determine the Modal class+ estimate the Mean, median, and range of a set of grouped data, selecting the statistic most appropriate to your line of inquiry or frequency graphs as long as it works. You might need to attach some of it so i can download it for Instance if you have done any graphs to show your working.

https://brainmass.com/statistics/confidence-interval/hypothesis-length-words-newspapers-3290

## SOLUTION This solution is FREE courtesy of BrainMass!

1. Null Hypothesis : That the mean Number of letters in a word in a newspaper articles is equal to the mean Number of letters in a word in a magazine. i.e the difference between the mean number of letters for each word is equal to 0.
<br>
<br>Alternative Hypothesis : That it is greater than or lesser that Zero.
<br>
<br>Step 1: Generate a set of 100 different samples for each category ( Newspaper and Magazine). It is done in the excel file attached along with. Formula Used = 10*rand()
<br>Assumption : Max No# of letters in a word can be 10 ( because rand()only gives uniform random variates between 0 and 1). Also Min No# of letters in a word can be 0 ( this cant be grammatically true , but will work just fine for statistical purposes).
<br>
<br>We will use the following Criteria for test purposes :
<br>a) if the difference between means is greater than 5 words at 95% confidence interval, reject the null Hypothesis.Else it holds True.
<br>
<br>Solution:
<br>a)Collect the mean of differences ( done in the excel sheet)
<br>B)Calculate the standard deviation of differences.
<br>
<br>Testing the Hypothesis:
<br>
<br>Sample Mean = 0 (This is Mean of Differences) ( shown in the excel book).
<br>Standard Deviation = 3.7984
<br>
<br>Range = {Mean - Z* Std Dev , Mean + Z* Std Dev}; Z=1.65 for 95% Confidence
<br> = {-6.067 , 6.42)
<br>
<br>Since the range on either side is greater than 5 words, reject null the hypothesis.
<br>
<br>Also only statistic of interest was the Means so no other statistic is calculated.
<br>The scatter plot with a linear line of regression is shown in the excel worksheet (along with the equation of the line).

This content was COPIED from BrainMass.com - View the original, and get the already-completed solution here!