Do you need help from SPSS to be able to test a hypothesis? Students are sometimes confused about the hypothesis, the meaning, and the numerical values in the SPSS output window. The goal of this SPSS tutorial is simply to help students understand the basic idea behind testing a hypothesis. Keep in mind that different projects and experiments will have very different testing parameters and criteria. Consult with your instructor or a qualified statistical expert if you have detailed questions about your particular project.
Before we can draw conclusions about a project, experiment, or study, we need a statement to prove whether it is “true” or not. Suppose we work for a major soft drink manufacturer. The company’s scientists have created a new formula for our most popular drink. Before spending millions promoting the new drink with TV commercials and endorsement deals, the company wants to know if people prefer the taste of the new formula over the old formula. How would the company do this?
The company could choose to carry out blind taste tests on a sample population, a population smaller than the entire country, for example, but with the same composition of ages, ethnicities, etc. The steps to obtain an accurate sample population are beyond the scope of this article. Since it would be impossible for everyone in the country to test the new formula, the company could conduct its experiments in test markets like New York or Los Angeles, which have a population that closely represents the population makeup of the entire country.
After identifying a sample population and conducting blind taste tests, the company would enter the results of its findings into SPSS. Before running a significance test, it would be necessary to identify a null hypothesis. The null hypothesis is the statement of no difference. Tea null hypothesis (H)0) represents a theory that has been put forward, either because it is believed to be true or because it will be used as the basis for an argument. It is a claim that has not been proven. In our example, the null hypothesis could be “People who drink our brand have no preference for the new soda formula over the old formula.” Whatever your experiment, the null hypothesis essentially states ‘There is no difference between A and B’. What we hope to do (especially before we spend millions of company money) is make the converse of the null hypothesis true.
What is the opposite of the statement “There is no preference for the new formula of soda versus the old formula”? The opposite of the null hypothesis is the alternative hypothesis (H1). Our alternate hypothesis could be stated as “People who drink our brand prefer the new formula to the old.” The alternative hypothesis for your particular experiment could be expressed as “There is a significant difference between A and B.” If there is a difference (and in our example we hope there is) how do we show it?
As with most statistical tests, we must be able to reject the possibility that the results we find are due to sheer luck. To do so, the test results must meet the significance level, or alpha, which sets the threshold for how extreme the data must be before rejecting the null hypothesis. Typically, alpha is set to zero point zero five. When you run a significance test, such as the paired-samples t-test or the independent-samples t-test in SPSS, the results window will display your results in a columnar table filled with numbers. You should see columns for the mean, standard deviation, mean standard error, degrees of freedom, and significance. You may have other columns and values depending on the particular test you are running. Let’s say we’ve run our blind taste tests, input our data, and run our significance test. (There are different types of significance tests. You must choose the correct one for your particular experiment, study, etc.). In our results window, under the significance heading, is the number zero point zero four two. What do all these different numbers mean?
To keep things simple, let’s focus our attention on a single value, the significance value. In a column labeled “sig,” you’ll find the significance value of your significance test. Take a look at the significance value. Is it greater than zero point zero five? If it is, it doesn’t matter. If the significance value is above the zero five point, you cannot reject the null hypothesis. In his report, he would indicate that he was unable to reject the null hypothesis. Sorry, but there is no difference between A and B.
Is the value less than zero point five? If so, there is meaning. In our example, the significance value is zero point zero four two. In this case, we can say that we reject the null hypothesis in favor of the alternative hypothesis. The results were not due to chance. There is a difference between the new soda formula and the old soda formula. There is a difference between A and B.
Special attention is paid to the null hypothesis. This is due to the fact that the null hypothesis relates to the statement being tested, while the alternative hypothesis relates to the statement that will be accepted if the null is rejected. The final conclusion, once the test has been carried out, is always given in terms of the null hypothesis. The result is “Reject the null hypothesis in favor of the alternative hypothesis” or “Cannot reject the null hypothesis”; the conclusion is never “Reject the alternative hypothesis” or “Accept the alternative hypothesis”. If the conclusion is “The null hypothesis cannot be rejected”, this does not necessarily mean that the null hypothesis is true. It only suggests that there is not enough evidence against H0 in favor of H1. Rejecting the null hypothesis suggests that the alternative hypothesis may be true.
The null hypothesis essentially states that the cases or items under consideration are statistically the same or exhibit the same behavior without any significant difference. The alternative hypothesis states that the given cases exhibit different behavior or that they have a statistically significant difference.
The two most important things to remember when working with a hypothesis test are the null hypothesis and the significance value in the output window. Remember that the null hypothesis is the statement of no difference. You should study the concept of alpha or significant value further until you know what to do when you get a number above the zero point five or below the zero point five that appears in the SPSS output window.
For an SPSS video tutorial on testing a hypothesis and tests of significance, such as the related samples t-test, visit the SPSS Help and Tutorials.