Theory Section part 2. Flashcards
Can skewness and kurtosis of a distribution be used to check normality? How?
If the distribution is perfectly normal, you would obtain a skewness and kurtosis value of 0
(Sjå også Theory Section Part 1, Skewness & Kurtosis)
Correlations analysis and regression both measure relationships between variables. What then is their difference?
Multiple Regression is a more sophisticated extension of correlation and is used when you want to explore the predictive ability of a set of independent variables on one continuous dependent measure. Different types of multiple regression allow you to compare the predictive ability of particular independent variables and to find out the best set of variables to predict a dependent variable.
Many statistical techniques assumes that the distribution of scores on the dependent variable is “normal”. What is normalty?
Normal is used to describe a symmetrical, belllshaped curve, which has the greatest frequency of scores in the middle with smaller frequencies towards the extrems
Normalty can be assessed (vurdere) to some extent by obtaining skewness and kurtosis
What is a “Null hypothesis”?
Null Hypothesis is a statement of the status quo (noverande/eksisterande tilstand), one of no difference or no effect. If the null hypothesis is not rejected, no changes will be made
H0: β = 0
H0: Det er ingen samanheng mellom Privacy Concerns og Willingness to buy online»
What is a “Alternative Hypothesis”?
Alternative Hypothesis is a statement that some difference or effect is expected. Accepting the alternative hypothesis will lead to changes in opinion or actions. The alternative hypothesis is the opposite of the null hypothesis
H1: β ≠0
H1: «Det er ein samanheng mellom Privacy Risk og Willingness to buy online»
Når bør ein teste to-sidig og når bør ein teste ein-sidig?
H1: β ≠0 Tosidig testing (Det er ein negativ eller positiv samanheng mellom Perceived Risk og Willingness To Buy Online)
Dersom ein ikkje har godt nok teoretisk grunnlag for det ein skal undersøke (teoretisk tungt forankra) skal ein teste tosidigt
H1: β > 0 (Det er ein positiv samanheng mellom Perceived Risk og Willingness to buy online)
H1: β < 0 (Det er ein negativ samanheng mellom Perceived Risk og Willingness to buy online)
What are the two main classifications of hypotheses?
- Tests Of Assosiation (H0: There is no assosiation/relationship between…)
- Tests Of Differences (H0: There there is no difference between…)
What “differences” can you test for (Test Of Differences)?
- Distributions (Compared to the population)
- Means (Between groups)
- Proportions (Expected vs. actual)
- Ranking (Median)
What “assosiation/relationship” can you test for (Test Of Assosiations)?
- Positive Assosiation (Correlation)
2. Negative Assosiation (Correlation)
Do you have any examples of Positive- & negative assosiation?
Positive assosiation: excitement -> purchase behaviour, shopping satisfaction -> re-patronage intention
Negative assosiation: satisfaction -> brand switching intention, perceived risk -> purchase intention
Do you have any examples of testing for differences?
“There is no difference between the two leading brands of natural water
There is no difference in levels of personal stress at work for males and for females ”
“The control group and the test market group of customers do not have different views of the brand image”
“Students’ attitudes to the brand were no different before than after viewing the TV advert”
What are some of the different statistics tools that can be used to test them (the two main classifications of hypotheses)?
Testing for differences: 1. T-test (Parametric, indep. sampl.) 2. ANOVA (Parametric, indp. sampl.) 3. Paired T-test (Parametric, paired sampl.) 4. Chi-Square (Non-param. indep. sampl.) 5. Mann-Whitney (Non-param. indep. sampl.) 6. Median (Non-param. indep. sampl.) 7. Kruskal-Wallis (Non-param. indep. sampl.) 8. Wilcoxon (Non-param. paired sampl.) Testing for assosiations: 1. Correlation analysis 2. Regression analysis
