Various methods such as t-tests and z-tests are used in testing a hypothesis within a given population. The use of a certain test method depends on particular conditions in existence such as the number of the sample. This paper discusses three familiar workplace situations where t-test can be used to test a hypothesis. It also proposes when to use a hypothesis test.
Where to Use T-test
Most organizations have their employees working in shifts of day and night, and since the small organizations such as tour operators have a small number of employees, t-test will be the ideal test compared to z-test to use in measuring the work output of their employees working in different shifts. A hypothesis in this case would be that day employees work output is similar to that of night shift employees.
Another situation in a workplace where t-test is applicable is in measuring the level of knowledge of new employees in a company. It is common for many organizations to hire employees in small numbers. Therefore, to measure the technical knowledge of new employees, t-test will be a better method than z-test. Similarly, t-test can be used in testing the daily customers’ satisfaction from their responses. Most organizations collect responses from a small sample of their customers using open-ended tools which make standard deviations to be unknown. This means that z-test cannot be applicable in this case.
When to use Hypothesis Test in Education
A hypothesis test is applicable in testing the differences in two population proportions in education since a learning setting provides a number of unique situations. For instance, the performance of boys against girls in an arithmetic class can be measured well by a hypothesis that boys normally perform better in arithmetic than girls. The data collection for the test would be possible since boys and girls are two distinct populations whose class performances are of normal distribution.
It can also be used when establishing the students’ performance based on their family backgrounds for a randomly selected group of student. Formulating a null hypothesis is simpler, and the responses in one group of student from a given background in this case are independent of the responses of the other. Finally, it can also be used when establishing the students’ performance in co curricula activities since some students like sports while others are anti sports. Getting distinct population samples is therefore easier and distributions of variables in this case are normal.
In conclusion, t-test is basically applicable when the variables being handled are small in size and in cases where the standard deviation is not known.