Hypothesis: Living on campus makes it easier to make friends. -

Hypothesis: Living on campus makes it easier to make friends.
Question11: Do you live on campus? Question66: How likely are you to share your work with fellow students?

In this questionnaire, we would like to understand whether the hypothesis holds or not so using the appropriate analytic method is necessary. The both two questions can be categorized into categorical data because the directions towards nominal and ordinal. Therefore, Chi-square test is suitable for testing the hypothesis. Moreover, we can adopt test of independence because it can be used to analysis whether the both variables interact, not the difference between these.

As we can see that there is not an noticeable relationship between the people live on campus and the level of sharing the work with fellow students. However, it might not be a good hypothesis due to the relationship between these two question are not relevant enough. Hence, I would not like to change the questions and the hypothesis, if I had the chance to do so. The first reason for this is that not the significant relationships need to be tested to ensure this hypothesis can be held, but the unexpected ones.

In this case, the independent variable is do you live on campus, which are identified in question eleven. The second variable is how likely are you to share your work with fellow student. As the level of sharing work with fellow students is affected by the first variable, it is the dependent variable. The values are given by the answers of question 66.

We can see here that Chi-square(1)= 6.915, P = 0.014. There is an unobvious phenomena of observed frequency and expected frequency because of P >0.05. This shows us that there is no statistically significant association between living on campus and the level of sharing work with fellow students. The values of both variables can be put into a logical order, raise with a stable growth and the 1.73 is meaningful. Thus, both variables have the same measurement level, namely ratio. Since the variables are not categorical (nominal) it is not relevant whether it is a within or between subject design.

Consequently, the widest range of tests is possible. If the data are parametric, even regressions (which are very powerful tests) can be run. This is the main reason, why I would not change my hypothesis. To see, if the data from the questions above are parametric, one needs to use descriptive statistics.

Chi-Square Tests
Value Df Asymp. Sig. (2-sided)
Pearson Chi-Square 6.915a 4 .140
Likelihood Ratio 7.321 4 .120
Linear-by-Linear Association .959 1 .328
N of Valid Cases 203
a. 1 cells (10.0%) have expected count less than 5. The minimum expected count is 1.79.

Result: The value is 6.915 and P is 0.14 which is >0.005, hence this hypothesis is not available. The question eleven which is do you live on campus and question sixty-six which is how likely are you to share your work with fellow students are not relevant to each other.

Cleared data:

This table below tells us to understand that mean is 1.73, minimum is 1 and the maximum is 2.

Descriptive Statistics
N Range Minimum Maximum Mean Std. Deviation Variance Skewness
Statistic Statistic Statistic Statistic Statistic Statistic Statistic Statistic Std. Error
Do you live on campus? 206 1 1 2 1.73 .443 .197 -1.061 .169
Valid N (listwise) 206
Potential outliers are above seven using the standard deviation rule (0.443 * 2 + 1.73 = 2.616). I decided not to conclude them as outliers, because the highest amount is one which is still a reasonable amount of living on campus. The dependent variable is ratio, but the skewness is below one. Thus, the data do not meet the criteria for parametric testing.
As can be seen the below table of how likely are you to share your work with fellow students, the minimum is 1 and the maximum is 5. Moreover, the Std. deviation is 0.990.

Latest Assignments