Search
Create
MR Final Zank
STUDY
Flashcards
Learn
Write
Spell
Test
PLAY
Match
Gravity
Terms in this set (37)
Casewise Deletion
Deleting the entire survey from one missed question.
-If you have a low number of respondents you don't want to do this.
-Unnecessary
Pairwise Deletion
Deleting only the variable that was not answered from the study.
-more common
Coding
the process of transforming raw data into symbols.
single variable
Coding: If one response is possible, what kind of variable?
-What is your opinion of Target?
(unfavorable _,_,_,_,_,_ favorable)
separate variable for each possible response, depending on number of questions.
Coding: If multiple sections under one question like a "check all that apply"
Factual
When coding open ended questions: Factual or Exploratory
- When there are multiple responses possible but only one correct response per person.
Exploratory
When coding open ended questions: Factual or Exploratory
-When there are multiple responses possible for a respondent with no one correct answer (opinion)
Steps to hypothesis testing
1. Formulate Ho and Ha
2. Select stat test
3. Choose alpha
4. collect data and test stats
5. Determine probability associated with test stat.
6. Compare with alpha
7. Reject or Fail to Reject Ho
8. Draw MR conclusion
Null Hypothesis
The status quo. No difference or effect expected.
Alternative hypothesis
Some difference or effect is expected.
strict equality, π
-Directional test.
-If inequalities are used.
-proportion or percent
π or µ?
two tailed test, µ
-If test is not expressed directionally.
-Equal sign is used.
-mean or average
π or µ?
Type one error, alpha
Reject the null hyp. but should not have. alpha or beta?
Type two error, beta
Fail to reject the null hyp. but should have rejected the null hyp. alpha or beta?
Level of significance
100%- confidence level
.15
If confidence level is 85% then alpha =?
Reject Ho
The larger the sample size the easier it is to?
=
If sig is ≥ alpha than ____ variance
≠
If sig is < alpha than ____ variance
Reject Ho and explain why.
When P value < alpha
Fail to reject null hypothesis
When P value is ≥ alpha
SSPS: single mean
-One-sample test
-Test value= Ho and Ha value
Independent t test
- Test from different populations to calculate the mean.
Paired t test
-If two observations are from the same respondent.
-can be before and after
Analysis of Variance (ANOVA)
-Testing hypotheses for 3 or more samples
- Null is all means are equal
-Alternate is at lease one is not equal
-examines variation between and within categories.
-Use post hoc ONLY when reject Ho
Chi-squared
Testing the association between 2 variables. NOT the mean.
Scatter Diagram
-Looks to see if there is a
linear trend
when plotting data
≠; =
Correlation _ Causation; Correlation _ Relationship.
- -1; 1
- Negative; positive
- more of a correlation there is.
-Correlation can range from __ to __.
-A negative number has a ___ correlation. A positive number has a ___ correlation.
The larger the number the ______.
Regression formula
y=a+B(x₁)
Formula meanings:
- Dependent variable
- y intercept
- independent variable slope
- level of independent variable
- Y= ______
- a= ______
- B=_______
- x= _____
Single mean test
one question. Interval or ratio
Independent T-test
-Interval or ratio
-nominal
Paired T test
two interval or ratio
Anova
-one interval or ratio
-3 or more nominal or ordinal (groups)
Cross tab (chi squared)
two nominal or ordinal
find the association btw
Correlation and regression
"the relationship" "correlation" btw
-two metric variables
-ind. variable and one or more dep. variable
;