Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
26 Cards in this Set
- Front
- Back
|
what does the null or statistical hypothesis state?
|
-no difference
-no association/no relationship |
|
|
alternative hypothesis
research hypothesis |
not what you will be testing, but in conclusion what you would state if null hypothesis is rejected.
research hypothesis is an informal prediction as to what you expect to find. |
|
|
directional hypothesis
|
gives actual direction as to what variable will be greater of small than the other variables
|
|
|
two-tail hypothesis
|
there will a change,but not sure in what direction
|
|
|
one-tail hypothesis
|
states what direction the change will happen
|
|
|
A good hypothesis is.....
|
-measurable and understandable
-states the relationship/difference between variables -is plausable upon literature and pilot data -IDV/DV is stated in hypothesis |
|
|
What is step 3 of the three step hypothesis testing?
|
selecting level of significance for your test
|
|
|
alpha level
|
the probability of incorrectly rejecting the null hypothesis.
-standard alpha level = .05 -type I error rate -"1 chance in 20 that you will concluded there is a difference when there isnt'." |
|
|
beta
|
-probability of incorrectly failing to reject the null hypothesis when in fact the alternative hypothesis is true (null is false)
-type II error -usually .2 (or 20%) -"1 in 5 chance of being wrong or finding NO difference between your groups when there really is a difference." |
|
|
step 4 of hypothesis testing
|
collection and analysis of data
|
|
|
step 5 of hypothesis testing
|
critereon for evaluating evidence
-compare calculated value to critical value in a table -most tests decision rules says reject null if calculated value is at least as large as the critical value -compare computer generated p-value to level of significance -reject null if p-value is equal or smaller than level of significance |
|
|
step 6 of hypothesis testing
|
decided to retain or discard null hypothesis
|
|
|
what is the only way to reduce the chance of beta or alpha error?
|
increase sample size
|
|
|
true or false
the smaller we specify the the alpha level, the larger the beta error will be? |
TRUE
|
|
|
step 7 of hypothesis testing
|
Strength of assosciation: determines practical significance
-r squared -eta-squared -omega-squared |
|
|
an example of effect size
|
Cohen's d
|
|
|
what does cohen's d do?
|
estimates effect size
-usually a number between 0 and 1 -.20 = small; .50=medium, .80=large |
|
|
what is power?
|
-likelihood of rejecting a false null
-the probability that difference could be detected (high power is .80 or above). -directly related to sample size |
|
|
list the 9 steps of hypothesis testing
|
1. Null hypothesis
2. Alternative hypothesis 3. specify alpha level 4. specify effect size (9-step) 5. specify desired level of power (9-step) 6. determine proper sample size (9-step) 7. Collect and analyze data 8. refer to criterion 9. retain or reject |
|
|
step 4 of 9
|
effect size
- |
|
|
step 5 of 9
|
Desired Power
-probability of detecting meaningful deviations from null -power ranges from 0-1 (beta error) -power of .90 gives beta error of .10 |
|
|
step 6 of 9
|
determine sample size
-found by using computer, formula or table |
|
|
benefits of the 9 step version
|
-knows the probability of making at type II error
-can determine number of subjects needed to get adquate power |
|
|
Confidence Intervals (CI) for hypothesis testing
|
-if null number outside CI reject null
-if null number contained in CI retain |
|
|
Inflated type I error
|
multiple tests conducted yields a probability of type I error being greater than that stipulated
|
|
|
what technique is used to deal with an inflated type I error?
|
bonferroni technique-divide type I error risk by number of tests-modifies experiment-wise error rate
|
