1&3 MINI ASSIGNMENT HELPFUL REMINDERS!! MUST READ!! 1 MINI ASSIGMENT REMINDERS Hypothesis

1&3 MINI ASSIGNMENT HELPFUL REMINDERS!! MUST READ!!
1 MINI ASSIGMENT REMINDERS
Hypothesis
• Null – there is no relationship between the two variables. For example, o There is no difference between men and women and the amount they are willing to pay for healthy food on campus.
• Alternative/Research – there is a relationship between the two variables. For example, o There is a difference between men and women and the amount they are willing to pay for healthy food on campus. (non-directional/two-tailed) o Men and willing to pay more for healthy food on campus than women are. (directional/one-tailed test)
• You either reject or fail to reject the null hypothesis.
Variables
• You are asked to convert an interval or ratio level variable into an ordinal level with no more than three levels.
• If you wanted to convert a Likert scale question you could do three categories (instead of the original five): 1. Strongly agree and somewhat agree 2. Neutral 3. Strongly disagree and somewhat disagree
• If you wanted to convert the age variable, you could do two categories (instead of the original five) 1. 18-23 2. 24+
Descriptive Statistics
• Do not include the descriptive statistics output. Choose the appropriate statistics from the output and present only those. Alpha/Probability Level/Terminology
• Remember to state the alpha – use .05 or .01
• Remember to report the p-value that was calculated. It is based on X 2 and the critical value but it’s important to show the reader the actual p-value that was calculated since it’s the probability of finding a value as extreme or more extreme than the X 2 that was calculated.
• There are arguments to be made of different alphas but for our purposes using one of the above is sufficient
• The results are never “statistically insignificant.” Instead, use the following terminology (adapt to suit your hypothesis):
o There is no statistically significant difference between….
o The difference between [the variables] is not statistically significant
Observed vs Expected
• Show the observed and expected values so the reader can see where the differences are. You will also interpret these for the reader, but it is helpful to have them in table format.
Degrees of Freedom
• To calculate degrees of freedom with chi-square you need to calculate the following o number of rows -1 * number of columns – 1
• For example,
o 3 rows and two columns would be 3-1*2-1 = 2 degrees of freedom
o 3 rows and three columns would be 3-1*3-1 = 4 degrees of freedom Interpretation
• Provide an interpretation of the results – what does it mean?
• Provide recommendations for the College. These should be more than a sentence.
o What should they do or not do based on this analysis?
o Make sure what you recommend relates back to the analysis you conducted. Don’t make statements that go beyond the data. If you think there is something else, they could be doing that is not related to the analysis you conducted then you could provide the recommendation that they do further investigation or conduct additional analysis and so on.
o Please be aware of the presentation of your analysis.
o Make sure your Word document is clear and easy to follow.
• Do not include the “descriptive statistics” output. It provides too much information that you are not interpreting for the reader.
3 MINI ASSIGMENT REMINDERS
Question 1
• It is not necessary to show the output in its entirety. Just pick out the key piece of information and present those.
• Do not show the whole dataset and do not submit an Excel spreadsheet.
• The t-critical that you use to compare the t-stat needs to be based on the type of alternative hypothesis you made – directional or non-directional.
• The question required you to draw conclusions so you should speculate why there was a difference between the amount of time spent watching TV based on marital status.
• You do not need to have equal observations in each group.
• Please use Excel to do the calculation.
• This should be a two-sample t-test and not a chi-square. You have ratio level data – use it!
Question 2
• Excel does not have a function to do a one-sample test…it’s rarely used in real life anyway. So, you do need to do part of the calculation by hand.
Question 3
• This is a paired-sample t-test so make sure you choose the appropriate test out of the list.
• A paired sample means that you have two observations for each subject. In this case, it is before and after. In a two-sample t-test (as in question 1) the two groups are independent.
• It is not necessary to show the output in its entirety. Just pick out the key piece of information and present those.
• You should use a directional hypothesis because you want to know if crime went down.
• Please use Excel to do the calculation.
• Don’t forget to offer recommendations to the Federal Government – what do the results mean?