Вы находитесь на странице: 1из 2

Hypothesis Testing

Hypothesis tests give you a way of using samples to test whether or not statistical claims are likely to be
true. Examine the Claim, Examine the Evidence, and Make a Decision

Six steps for hypothesis testing:

1. Decide on the hypothesis you’re going to test

2. Choose your test statistic

3. Determine the critical region for your decision

4. Find the p-value of the test statistic

5. See whether the sample result is within the critical region

6. Make your decision

1. Decide on Null and Alternative Hypothesis (H0 - default position and claim accepted unless there is
strong evidence against it (counterclaim HA). There is also where Null similar to "no," and states there is
"no relationships" between variables being tested. N0 and NA are opposite. A good research questions is
a clear focused and concise question that drives the study it contains variable and relationships being
tested. This can be HA and then formulate H0.

2. Choose your test statistic (the statistic that’s most relevant to the test. The test statistic is used to
work out the probabilities used as evidence).

- Discrete
 -Geometric Distribution
 -Binomial Distribution
 -Poisson Distribution

- Continuous

 -Gaussian/Normal - >Standardized Distribution


 -t-distribution -> Standardized T

2a. Draw out the distribution with H0, HA

3. Find the Critical Region. The critical region of a hypothesis test is the set of values that present the
most extreme evidence against the null hypothesis. To find the critical region, first decide on the
significance level α (usually 0.1(z-value = ± 1.28), 0.05(z-value = ± 1.645), 0.01(z-value = ± 2.33)).
P(X < c) < α.

 NOTE: Other books call "c" the confidence level. c and α are complementary, i.e. c + α = 1. If a
problem give either c or α, the other can be calculated (e.g. if α=0.05, then c=1-α =1-0.05 = 0.95)
 NOTE: The most common significance level is 5%, although you sometimes see tests at the 1%
level. Testing at the 1% means that you require stronger evidence than if you test at the 5%
level.
 A one-tailed test is where the critical region falls at one end of the possible set of values in your
test.
 A two-tailed test is where the critical region is split over both ends of the set of values. Both
ends contain α/2, so that the total is α. You can tell if you need to use a two-tailed test by
looking at the alternate hypothesis HA. If HA contains a ≠ sign, then you need to use a two-
tailed test as you are looking for some change in the parameter, rather than an increase or
decrease.

3a. Draw out the Critical Region(s)

4. Find the p-value. This is the calculated probability, of obtaining a result equal to or "more extreme"
than what was observed. How to find the p-value depends on the distribution one is using- calculate the
probability using the distribution and their parameters.

5. Is the sample result in the critical region? Compare α to p-value.

6. Make your decision. Accept or reject the Null Hypothesis (H0) (e.g. for a lower one-tailed test, is p-
value < α, if so, reject H0, because the results are statistically significant, as they’re unlikely to have
happened by chance) and therefore accept/reject HA.

Вам также может понравиться