Embedded Files
As an hypothesis, we have decided to test whether or not female consumers take into consideration the amount of calories being ingested when consuming beer.
We are consucting a hypothsis test as its objective is to determine if there is enough statistical evidence in favor of a belief or hypothesis related to a parameter.
Normality and Independent Testing
1.Hypothesis:
H0 : Females take into consideration the calories of beer when consuming.
H1: Females do not take into consideration the calories of beer when consuming.
[spss output 26 - tests of normality]
As the dimension of the samples for both males and females are more than 50 in the normality testing, we consider and use the Kolmogorov-Smirnov Test.
We note that the p-value is inferior to the assumed significance level of 0.5. Furthermore, as a consequence, we reject the null hypothesis (H0) and assume there isn't any normal distribution in this test of normality.
[spss output 27 - independent sample test]
Moreover, as we have to reject the null hypothesis, in the Levene's Test for equality of variences we cannot assume equal variences for gender and caloric intake.
Wilcoxon Signed-Rank Test
2.Hypothesis:
H0 : The median of differences between drinking beer helps one relax and beer tastes better with friends equals 0.
H1: The median of differences between drinking beer helps one relax and beer tastes better with friends does not equal 0.
We are choosing this hypothesis to analyze the fourteenth question that describes the levels of agreement that consumers rated based on the sentences that best decribed the situations related to beer consumption and its effects.
We chose to test for:
14.1 - "Drinking beer helps me feel good."
14.3 - "Beer tastes better if you drink it with friends."
[spss output 28 - hypothesis test summary]
[spss output 29 - related-samples Wilcoson sighed rank test summary]
When performing the Wilcoxon signed-rank test, there is a significance level of 0.05 that is being considered. The p-value shows to be inferior to the significance level and so we have to reject the null hypothesis and assume the two distributions aren't the same for each of the sentences analyzed in the ranking.
By looking at the graph, we can see that the difference was set as Beer tastes better if drank with friends minus Drinking beer helps one relax.
There are 72 positive differences, 16 negative differences and 38 ties.
With this, we can say that the Beer tastes better if drank with friends statement is the one consumers agree with the most as we note more positive differences than negative ones.
[spss output 30 - related-samples Wilcoxon signed rank test]
[spss output 31 - categorical field information drinking beer helps one relax]
Bar chart representing the level of agreement/disagreement of consumers on if drinking beer helps one to relax
In the graph that ranks the statement "Drinking beer helps one relax," we see that there was a high amount in consumers that agreed to it. However, in comparison to 'strongly agree', there were more consumers that felt 'undecided' about this statement.
On the other hand, in the graph that ranks the statement "Beer tastes better if drank with friends", the highest level of agreement was 'strongly agree.' This was followed by 'agree', and showed low rankings on the disagreance of the statement. We believe it is important to also highlight that in this graph there was no ranking in the category of 'strongly disagree', which we believe is of great importance and portrays consumers identify themselves mostly with this statement.
[spss output 32 - Categorical field informatio beer tastes better if drank wi friends]
Bar chart representing agreement/disagreement of consumers on if beer tastes better if drink with friends
Chi-Square Test
3.Hypothesis:
H0 : Frequency of consumption is independent of district of residence.
H1: Frequency of consumption is not independent of district of residence.
[spss output 33 - Chi-square test]
We make an observation to the percentage of the cell and note a 87.5% value with the expected count less than 5. As we have a value higher than 20%, we cannot assume the validity of results ran in the test.
For the analysis of this Chi-Square test, we consider the first row, it being the first value for the asymptotic significance (2-sided).
The p-value observed in this test is 0.020. When taking a level of significance of 0.5, we reject the null hypothesis (H0).
The frequency of consumption is dependent of the district of residence consumers are situated in.
[spss output 34 - Frequency of consumption crosstabulations]
As noticed in the crosstabulation, 65.5% of consumers that reside in Aveiro drink beer once or twice a week. However, 100% of people that reside in Braga and Castelo Branco are rare drinkers.
Furthermore, because 13.3% of consumers reside in Leiria, they consume beer once a day. In Lisbon, residents make up 50% of the rare drinkers, and 100% of residents in Viana do Castelo drink beer once or twice a week.
48.6% of consumers in Porto drink once or twice a week, and we thought it would be important to analyze this that depending on the district a consumer is located in, there is also an impact in influence when regarding consumption of beer, even if its just the slightest, there is still significance to it. As evidence, we note that because consumers live in Viseu, 3.3% of the consumption is once or twice a week.
Page updated
Google Sites
Report abuse