A null hypothesis is a hypothesis that says there is no
statistical significance between the two variables. It is usually
the hypothesis a researcher or experimenter will try to disprove or
discredit. An alternative hypothesis is one that states there is a
statistically significant relationship between two variables
Hypothesis Testing for Small Sample Quantitative Data
Learning Objective:
Hypothesis Test
Step 1: Make assumptions about your data
- For these tests, we assume that our data is quantitative and
that the population is normally distributed. We also know
that our sample size is going to be relatively small, which means
that we will be referencing the t table rather than the z
table.
Step 2: Formulate null and alternative hypotheses
- Just like with other types of hypothesis testing, the null is
what a determined skeptic would believe about what we are going to
measure, and the alternative hypothesis is our research hypothesis
based on the data collected.
- As with other types of hypothesis testing, this test can be
either one-sided or two-sided. Make sure you keep in mind how you
intend to test the data.
Step 3: Calculate a Test Statistic
- This statistic summarizes how much our alternative hypothesis
differs from the rest of the data if the null was true. Here is the
formula, which we’ve used many times before:
- Test statistic = ( Xbar -)/(s/sqrt(n))
- Don’t forget to note the degrees of freedom!
- d.f.= n-1
Step 4: Calculate a p-value
- Remember, a p-value is a measure of surprise, so small p-values
more strongly contradict the null because that means it would be
extremely surprising to see the results of our alternative
hypothesis if the null were true.
- To find the p-value, go to this t table or we can use R.
Keep in mind the confidence interval, the degrees of freedom, and
if the hypothesis is one or two-sided.
Step 5: Draw a conclusion
- Are the results statistically significant given the
pre-determined
level
- If p value <
we reject the null and conclude that the evidence supports the
alternative hypothesis. If p- value >
.we fail to reject the null.