This paper highlights the type of data that supports the use of t-test for statistical analysis. The choice of statistical analysis is often made before carrying out an experiment. For proper analysis, the choice of effective static is of importance. It therefore begins with the understanding of the type of data. Data is either continuous or discrete and categorical (Kachigan, 1991).
T-test is used in data type that involves the comparison of values of the mean from two samples. It is used in such types of data that needs determination of whether the mean of two groups are statistically different from one another. Simply put, a t-test is used for the analysis of data type that needs comparison of two or more groups.
A t-test analysis is used with data type that have values in one group that appear to be more closely correlated with a specific value in a different group than with random values in the other group. However, when the data collected is matched or paired, a paired t-test should be selected for analysis (Kachigan, 1991).
It is also used when the data needs statistical hypothetical test. T-test is used with type of data that follows normal distribution. Besides, this statistical analysis technique is used when the data is in two sets which are form of independent and identically distributed samples.
A t-test should also be used when the data collected is of paired samples. In this case, dependent t-test is employed for analysis. It appropriate where the data has been matched or paired. On the whole, the data should offer mean of two normal populations with unknown but same variances (Kachigan, 1991).
In health services, for example, a t-distribution can be used to compare the mean sampled to some constant known value. A researcher may be interested in determining whether the number of cancer patients in Sheffield city residents was higher or lower than the average disease prevalence in the entire UK. Suppose the average cancer prevalence in UK is 35,000 patients per year, mean cancer prevalence from a random sample of Sheffield can be determined using a one sample t-test.
Reference
Kachigan, S. K. (1991). Multivariate Statistical Analysis: A Conceptual Introduction. 2nd. Radius Press.
