Question

Use the following problem, data set, and SPSS outputs to answer questions 16-21 You conduct a study to examine whether the beverage a person drinks affects their willingness to socialize. You select 21 people at the mall and randomly give them Water, Coffee, or Gasoline. 10 minutes after they drink you ask them to rank their willingness to socialize on a scale of 1-10. You predict that compared to Water drinking Coffee will make people feel more sociable and that drinking gasoline will not affect their rating (compared to water). For this experiment you decide a = 0.05. You conduct a One-Way ANOVA in SPSS comparing Willingness to Talk by Condition (Water, Coffee, Gasoline). The output from that analysis of are as follows: ited States)

Answer 1

16) We Use ANOVA when we want to test effect of independent variable on dependent Variable but the level of that independent variable should be three or more . Here our dependent Variable is socializing willingness and independent Variable is type of drink that person consumes. There are three levels of one independent variable for which we are testing if average willingness to socialize is same for all three drinks or it differs .

This can be tested Using One way ANOVA .

17) Null hypothesis : Average willingness to socialize is same for all drinks.

Alternative hypothesis : Average willingness differs atleast among one pair.  

18) There are key parameters to look at in the generated output . F statistic for drink is 18.323 p-value ( sig.) Is 0.000 . These values when compared to predetermined level of significance will help us to take decision regarding acceptance or rejection of null hypothesis. Multiple Comparison table gives values to compare a drink with other drinks.

19) Level of significance = 0.05 , p-value = 0.000

Since p- value is less than level of significance we Reject Null hypothesis can conclude that there is difference between level of willingness to socialize for three drinks.

20) Yes it is necessary to Conduct Post Hoc tests since we have Rejected Null hypothesis. Alternative hypothesis says that there is difference but it doesn't tell you between which drinks is the difference. So to find the pairs we need to do post hoc analysis.

21) Looking at the results of Post Hoc analysis we can see that there is difference between water and Coffee ( if p-value is less than 0.05 means difference exists)

There is difference between coffee and gasoline and between water and gasoline.

So finally we can say that level of willingness to socialize is differing on consumption drinks.