Chapter Seven

Decision Table, P Value less than Alpha

Reject the null, statistically significan, may be clinically significant if the experts say, find a relationship. Correct or Type one (chance only)

Effect size/ Sample Size

as effect size increases, sample size decreases

Power size/ sample size

As power increases, sample size is increased, and the likelihood of rejecting the null hypothesis correctly is increased. The larger the sample size the greater the power of the study (more likely to accurately reject the null hypothesis)

Ho True, Accept Ho, Fail to Reject the Null

Correct acceptance of Null

Decision Table, P Value greater than Alpha

Fail to reject the null, not statistically significant not clinically significant, no relationship. Correct to Type Two (usually too small of sample, power error)

Power Analysis

How sample sizes are calculated. Determines what sample size will ensure a high probability that we correctly reject the null that there is no difference between groups. Pick power (usually 0.8), threshold for significance (usually 0.05), and then effect size which is the overlap

Power

The ability to find a difference or an association when one actually exists The likelihood of rejecting the null hypothesis correctly; that is you say there is a relationship and difference and you are correct

Alpha

The chance of making a type one error

Beta

The chance of making a type two error

Type Two Error

The error made when a researcher accepts the null incorrectly, missing an association that is really there (sometimes called a power error because the researcher may not have enough power to find an association that really exists). Ho False. Usually occurs because sample was not large enough

Type one Error

The error when a researcher incorrectly rejects the null hypothesis, when he or she concludes there is a significant relationship but there really is not. Ho true

Ho False, Accept Ho, Fail to Reject the Null

Type Two Error

Ho True, Reject the Null

Type one Error= Alpha

Example of Type One Error and Type Two Error

Type one error (false positive): to a male, saying they are pregnant Type two error (false negative): to a pregnant female, saying you're not pregnant

Type Two Power

You will either correctly identify (i.e, reject Ho correctly = power) or you will incorrectly miss it (type two error) Power (correctly rejecting) + Beta (chance of making a type two error) = 100% (Therefore if there is a 80% chance that you are correct and find the relationship (power), the chance of a type two error is 20%)

If The Null is not True

there is a relationship between the variables you can conclude one of two things: You can fail to reject it. In this case you are incorrect and are making a type two error. The probability of doing so is equal to Beta (usually 0.20 or 20%) You can reject it correctly. The probability of reaching this conclusion is 1- B (usually 80%), which is also the power of your study

Effect Size Numerical

weak is less than 0.3, moderate is 0.3 to 0.5, strong is 0.5

Ho False, Reject the Null

correct rejection of null= power= 1- Beta

Effect Size

The extent to which a difference or relationship exists between variables in a population (the size of the difference you are attempting to find)

Sample Size

an adequate sample size is largely determined by the size of the difference between group means within the population (effect size), you are attempting to find and the power needed to accurately find it

If the Null Is True

there is no relationship or difference you can conclude one of two things: You can fail to reject it. You are correct in this situation, and the probability of reaching this conclusion is 1 - alpha (usually 0.05) or 95% Or you can reject it. Then you are incorrect and are making a type one error. The probability of reaching this conclusion is equal to alpha, usually 5%