# Please explain to me in laymans terms how to calculate a single T-test

Hi,

Can someone explain to me in laymans terms how to calculate a single T-test?

Thanks,

Are you talking about a one-tailed T-Test? What exactly are you trying to accomplish, and what data do you have?

Some of the stat gurus may be able to provide some assistance. Here is a pretty decent step-by-step guideline (about 1/3 down the page) for non-stat gurus like myself. You might want to look at this, and then determine which parts you don’t understand.

http://www.microbiologybytes.com/maths/1011-20.html

Calculating a single sample T test consists of the following steps:

1. Check the data to confirm the data forms a normal distribution (approximately). This can be done by plotting a histogram and visually determining the data is an approximate normal distribution. Overlaying the histogram with a normal distribution curve is helpful for this determination. If you cannot tell visually, run a normality test. The T test is somewhat robust to normality, but if you get into distributions that are clearly not normal, you may have to use a nonparametric alternative analysis.

2. Determine the significance level (alpha) that you want and the comparison mean value that you want to test against.

3. Calculate the mean and standard deviation of your data.

4. Calculate the Test Statistic using the following method:

Test Statistic = (Data mean – Comparison mean)/((standard deviation)/Square root (number of samples))

1. For testing the Null Hypothesis: Data Mean = Comparison Mean, Calculate the lower critical value by LCV = - Table E value for alpha, degrees of freedom (N-1), and Upper Critical Value by UCV = Table E value for alpha, degrees of freedom (N-1).

2. Evaluate Null Hypothesis by comparing the Test Statistic to the LCV and UCV. If the Test Statistic is < LCV or > UCV, reject the Null Hypothesis in favor of the Alternate Hypothesis (Data Mean is NOT = Comparison Mean). Otherwise, fail to reject the Null Hypothesis, there is insufficient evidence to show that the means are different.

3. For testing the Null Hypothesis: Data Mean = Comparison Mean, Calculate the critical value = - Table D value for (alpha x 2), degrees of freedom (N-1).

4. Evaluate Null Hypothesis by comparing the Test Statistic to the Critical Value. If the Test Statistic is < Critical Value, reject the Null Hypothesis in favor of the Alternate Hypothesis (Data Mean is < Comparison Mean). Otherwise, fail to reject the Null Hypothesis, there is insufficient evidence to show that the data mean is < the comparison mean.

5. For testing the Null Hypothesis: Data Mean = Comparison Mean, Calculate the critical value = Table D value for (alpha x 2), degrees of freedom (N-1).

6. Evaluate Null Hypothesis by comparing the Test Statistic to the Critical Value. If the Test Statistic is > Critical Value, reject the Null Hypothesis in favor of the Alternate Hypothesis (Data Mean is > Comparison Mean). Otherwise, fail to reject the Null Hypothesis, there is insufficient evidence to show that the data mean is > the comparison mean.

I hope this helps.