What are the four steps of hypothesis testing?
(1) Null hypothesis
State the null hypothesis
(2) Set the criteria (Alpha Level)
(3) Collect data and compute sample statistics (z=(samplemean - hypothesizedpopmean)/standarderrorUandM
(4) Make decision whether the sample data has reached beyond the criti
For hypothesis tests how do you try to find a hypothesized population U?
In the z-score formula plug everything else assumption and if M-U equals 0 then you have found the population mean.
What are the zscore
(a) ?=0.5
(b) 0.01
(c) 0.001
(a) 1.96
(b) 2.58
(c) 3.3
What are the alpha levels for
(a) 1.96
(b) 2.58
(c) 3.3
(a) ?=0.5
(b) 0.01
(c) 0.001
A study concluded that the use of mindfulness helps resolve ADHD though the sample selected was East Asian. What type of statistical error could occur here?
Type 1 error where the null hypothesis is rejected when it is actually the sample that is very extreme.
A study investigating the effect on your attractiveness on Tinder when you post a shirtless photo of yours abs finds that out of a sample of n=6 profiles the null hypothesis could not be rejected. What type of statistical error could occur here?
In this case it is possible that the null hypothesis should be rejected but the test was not sensitive enough to detect it.
If you are an RA asked to draft a research paper for the at a about the birth rates of rats and the effect of alcohol in the mother and the data was z=3 thus defeating H0 then how would you state the results
The study with alcohol had a significant effect on the birth weight of the new born rats z=3, p<0.05.
What are three common assumptions made with Hypothesis test and Z-score?
(1) Random selection
(2) Not consistent observation between 1 and 2
(3) SD is not changed for a population before and after treatment but rather a constant is added.
What are the arguments for and against the one tail hypothesis test?
For: Is more sensitive to treatment
Against: Lowers the expectations
What is three major limitations of what a hypothesis test can tell us and what measure is used to try to make up for it?
(1) Hyp test focuses on data rather than on the hypothesis
(2) Significant effect does not always equal substantial
(3) Given enough researchers there will always be the probability that some samples commit type 1 error leading to a waste of resources whe
What does the effect size tell you? What is the equation? How much would be considered a small, medium, and large effect size?
You get the Cohen's D provides a a measure in terms of SD and identifies the distance at which exists between the untreated and treated population.
The equation for estimated Cohn's D = (Utreatment - Unotreatment)/SD
d=0.2 Small effect
d=0.5 medium effect
What is statistical power and how do you use it?
Statistical power is the likelihood a study will commit Type 2 error. Based on the concept that 1 - ? should be 0.
Step 1: Get Om
Step 2: multiply ?-Zscore by 2 (Only if one tail)
Step 3: Add M + Step 2 result
Step 4: Use step 3 as M and apply to Z-score
How does the following variables effect
(a) Sample
(b) alpha level change
(c) 2 tail to 1 tail
(a) Smalle n smaller O, and a large B
(b) Making the alpha more precise lowers the hypothesis
(c) Going from two tailed to 1 trailed inceases the power of the test.