Statistical Tests - Essay Example

Statistical Tests due: Introduction A bivariate analysis is a method of data comparisoninvolving exactly two measurements being made on observations. Another name for this method is inferential statistics. The two measurements can be called X and Since the two, X and Y are obtained for each observation, and then it means the data for each observation is the pair X and Y.A bivariate data can be stored in a table with two columns, as suggested by the term bivariate which means two variables. Bivariate analysis is the analysis of the relationship between one independent (causal) variable and one dependent (outcome) variable.

Different types of statistical tests are used differently depending on the situation (Jekel, 2007). Situation 1 Comparison is made between campaign contributions (in dollars) to a marketing campaign for healthy living for men and the contributions for women in the same campaign. A Chi-square will be used in this particular case. Chi Square compares the observed frequencies to expected frequencies, by making use of a t-Test. This test looks at differences between two groups on some variable of interest.

In this case, two groups are considered; that is the male against a female, in terms of their contribution towards a healthy living campaign. The variable (causal) is being measured against the variable (outcome) which is the contribution in dollars. The two causals are, therefore, being measured against the outcomes. Analysis of the two will give clear hypotheses on which gender is active in contributing towards the healthy living campaign, for a given period, which could be a month or year (Jekel, 2007).

Situation 2 Comparing the contributions (in dollars) to the campaigns for healthy living people in the age brackets 20-30, 30-40, and 40-60 years old. In this case, comparison is being made between causals who are the people in three different age groups against, the outcomes that are the contributions. The most suitable test to use in this case is the t-Test, just like in situation one above. This is because the variables in question are also two. That is the age is being compared with the contributions made for a given period (Timm, 2002).

Situation 3 Comparing the responses given by the HR and CEO, where they give the answers, yes, no or not sure to an attitude question. The best test for use in this case is anova. Anova checks the importance of group variations existing between two or more groups. It determines the difference between groups, but does not tell which is different. In this particular case, two groups of people are questioned, though no distinction is given on the genders. Also, the answers expected are not more than two.

The causals are the human resource managers and the chief executive officers while the expected outcomes are the responses given which could be “yes”, “no” or “not sure” (Jekel, 2007). Situation 4 In a mail survey, half of a sample received an incentive and the other have did not. Response rates are compared. A t-Test would be the best for this situation. This will be made possible by the causal variables involved. There are two equal samples being measured against their response time.

Differences are being investigated between the two groups, based on a variable of interest which in this case is the response time. Situation 5 Married men will push a grocery cart when they go shopping with their wives. The most suitable test in this situation would be ANCOVA. This test unlike the Anova test has additional features in that it has extra covariates influencing dependent variable (DV).The DV here is the pushing of a grocery cart by married men. However, this variable will depend on another factor, which is the wife to the man.

In some cases, the men might be willing to push the grocery cart but the wives may object. This means that this hypothesis may not apply to all married men. However, to test the hypothesis, Anova would add the covariate of the wives, thus proving to be the best (Timm, 2002). Situation 6 Comparing the job performance of a salesperson before and after undergoing ethics training. MANCOVA test would be best to use in this case. This is because it has additional covariates that influence the dependent variable.

In this case, comparison is being made between the performance of the salesperson before and after the ethics training. The variables are the performance against training. The performance, however, will be based on the dependent variable; ethics training (Timm, 2002). Conclusion To reach on a certain hypothesis, different statistical test method has to be employed as it has been discussed. The choice of a certain test will depend on the particular variables used in the study. Proper analysis of the given variables is done to obtain proper interpretation of a given situation.

References Jekel, J. F. (2007). Epidemiology, biostatistics, and preventive medicine review. Philadelphia, PA: Saunders. Timm, N. H. (2002). Applied multivariate analysis. New York: Springer.