# The Fundamental Concepts Of Statistics

1309 Words6 Pages
Question 5 This paper will provide a sample of the fundamental concepts in statistics. During research there is usually variance between individuals within a group or between different groups. During hypothesis testing the researcher wants to know if the sample of data collected is truly representative of the entire population. Null hypothesis testing is concerned with the correlation or differences in means (continuous data) between groups. The null hypothesis works on the premise that there is no difference between two groups such as males and females who respond to a set of items (independent sample T-test). The null hypothesis can also be described as the hypothesized mean being equal to the population mean. So the null hypothesis is…show more content…
Whereas, if I fail to reject the null, because I believe the groups are the same when, in fact, they are different this is a type 2 error. In its simplest terms, Type I error is the incorrect rejection of the null hypothesis, whereas Type II is the incorrect acceptance of the null hypothesis. Related to hypothesis testing is the Significance level or cut off point (alpha). The P or the probability value is one of the indicators that can be used to test the null hypothesis. Usually a significance level = .05 (95% confidence interval) or lower is the accepted rule for rejecting the null hypothesis. This means that we will accept up to 5% of type 1 error. Also, P < 0.01 sometimes is referred to as statistically highly significant. If the p value is < .05, then we would reject the null hypothesis. Visually, on a bell curve (two-tailed test) this can be presented as the critical region or how far our sample statistic is from the null hypothesis. In relation to P values there is the confidence interval (or bound). A 0.05 level of significance would be centered on a 95% confidence interval. If our given value sits outside the bound we would reject the null. However, if our value sits within the interval we would not reject the null hypothesis. Another concern during research is the issue of sampling error due to only a portion of the population being used during a study. Even when bias is reduced by random sampling the