Lecture week 3
- What
An estimate around a mean
- Why do we need to use the student-t distribution for a one-sample t-test?
We use an estimate of the standard deviation in the population
- What is the purpose of the one-sample t-test?
To test a claim about the value of an average
- When we find a p-value smaller than 5%...
we reject the null-hypothesis
- We do a two-tailed test when…
The alternative hypothesis is formulated as ‘not-equal-to’
- We do a one-tailed test when…
the alternative hypothesis is formulated as ‘larger than’ or ‘smaller than’
- How do we formulate the null-hypothesis in a Chi-square test?
There is no difference between the groups – the two variables are independent
- How do we formulate the alternative hypothesis in a Chi-square test?
There is a difference between the groups – the two variables are dependent
- What are conditions for the Chi-square test?
Not more than 20% of the expected frequencies are smaller than 5
All expected frequencies are larger than zero
- What does the Chi-square represent?
The size of the difference between observed and expected frequencies.
Lecture week 5
- What does a correlation of -0.67 between distance and shopping tell you?
The further one (lives), the fewer one (shops) (Much vs. little)
- If the correlation is 0, can you always conclude that there is no relationship?
no there may be a non-linear relationship
- For correlation analysis what level of measurement is required?
Both variables need to be ordinal or ratio
- True or false; ‘correlation requires determining which variable is the dependent
and which is the independent variable?
False ( True; if it is regression analysis)
- True of false; With correlation analysis one can predict the value of Y
(dependent) if one know the value of X (independent)?
False (only states the strength and direction of the relationship)
- What is or are the main goals of regression analysis?
Both prediction and explanation
- Given is the regression model; visits = 7.13 – 0.56 * distance. What is a right
interpretation?
Both answers are correct
(When distance = 0, the average number of visits = 7.13)
(With every extra kilometer the visits decrease with 0.56)
- What is an assumption of regression analysis?
Y has a normal distribution for fixed values of X
, Lecture week 6
- What is the unique goal of multiple regression analysis?
To identify the pure effects of the predictors
- In the context of regression analysis, what do we mean by ‘controlling other
variable’?
The effect of a variable keeping all other variables constant / fixed
- Standardized coefficient are useful especially for….?
Explanation
- What does the Stepwise method involve?
A automated stepwise manner of adding variables to the model
- When does multicollinearity occur?
When strong correlations exists between the predictors mutually
- Is it allowed to use a nominal variable as a predictor in a regression analysis?
Yes, provided that dummy variable are used
- When two predictors variables are strongly correlated with each other
(multicollinearity) what would be the best thing to do?
Both are good options:
1. Combine the two variables in one new variable
2. use only one of the two variables (remove one of the two variables)
Lecture week 7
Question 9. A manager of an office building is interested in the strength of the relationship
between two variables: the amount of energy used for heating (in kilowatt) in a work space
and the feeling of comfort in the work space. For a sample of 20 work spaces the energy use
and feeling of comfort are determined. The feeling of comfort is determined by an expert: the
expert judges the comfort by rank ordering the 20 work places from highest to lowest level of
comfort.
Which test should be used to investigate the strength of the relationship between these two
variables?
a. Spearman correlation
b. ANOVA
c. Pearson correlation
d. Kruskal-Wallis H test
Question 10: Which statement about non-parametric tests is NOT true?
a. A non-parametric test has more power than a parametric test for finding a difference or a
relation between two variables
b. A non-parametric test does not require that the variables are normally distributed
c. A non-parametric test is also safe to use when the number of observations is small
(smaller than 30, but larger than 10)
d. A non-parametric test can be used when one or both of the variables is of ordinal level
- What
An estimate around a mean
- Why do we need to use the student-t distribution for a one-sample t-test?
We use an estimate of the standard deviation in the population
- What is the purpose of the one-sample t-test?
To test a claim about the value of an average
- When we find a p-value smaller than 5%...
we reject the null-hypothesis
- We do a two-tailed test when…
The alternative hypothesis is formulated as ‘not-equal-to’
- We do a one-tailed test when…
the alternative hypothesis is formulated as ‘larger than’ or ‘smaller than’
- How do we formulate the null-hypothesis in a Chi-square test?
There is no difference between the groups – the two variables are independent
- How do we formulate the alternative hypothesis in a Chi-square test?
There is a difference between the groups – the two variables are dependent
- What are conditions for the Chi-square test?
Not more than 20% of the expected frequencies are smaller than 5
All expected frequencies are larger than zero
- What does the Chi-square represent?
The size of the difference between observed and expected frequencies.
Lecture week 5
- What does a correlation of -0.67 between distance and shopping tell you?
The further one (lives), the fewer one (shops) (Much vs. little)
- If the correlation is 0, can you always conclude that there is no relationship?
no there may be a non-linear relationship
- For correlation analysis what level of measurement is required?
Both variables need to be ordinal or ratio
- True or false; ‘correlation requires determining which variable is the dependent
and which is the independent variable?
False ( True; if it is regression analysis)
- True of false; With correlation analysis one can predict the value of Y
(dependent) if one know the value of X (independent)?
False (only states the strength and direction of the relationship)
- What is or are the main goals of regression analysis?
Both prediction and explanation
- Given is the regression model; visits = 7.13 – 0.56 * distance. What is a right
interpretation?
Both answers are correct
(When distance = 0, the average number of visits = 7.13)
(With every extra kilometer the visits decrease with 0.56)
- What is an assumption of regression analysis?
Y has a normal distribution for fixed values of X
, Lecture week 6
- What is the unique goal of multiple regression analysis?
To identify the pure effects of the predictors
- In the context of regression analysis, what do we mean by ‘controlling other
variable’?
The effect of a variable keeping all other variables constant / fixed
- Standardized coefficient are useful especially for….?
Explanation
- What does the Stepwise method involve?
A automated stepwise manner of adding variables to the model
- When does multicollinearity occur?
When strong correlations exists between the predictors mutually
- Is it allowed to use a nominal variable as a predictor in a regression analysis?
Yes, provided that dummy variable are used
- When two predictors variables are strongly correlated with each other
(multicollinearity) what would be the best thing to do?
Both are good options:
1. Combine the two variables in one new variable
2. use only one of the two variables (remove one of the two variables)
Lecture week 7
Question 9. A manager of an office building is interested in the strength of the relationship
between two variables: the amount of energy used for heating (in kilowatt) in a work space
and the feeling of comfort in the work space. For a sample of 20 work spaces the energy use
and feeling of comfort are determined. The feeling of comfort is determined by an expert: the
expert judges the comfort by rank ordering the 20 work places from highest to lowest level of
comfort.
Which test should be used to investigate the strength of the relationship between these two
variables?
a. Spearman correlation
b. ANOVA
c. Pearson correlation
d. Kruskal-Wallis H test
Question 10: Which statement about non-parametric tests is NOT true?
a. A non-parametric test has more power than a parametric test for finding a difference or a
relation between two variables
b. A non-parametric test does not require that the variables are normally distributed
c. A non-parametric test is also safe to use when the number of observations is small
(smaller than 30, but larger than 10)
d. A non-parametric test can be used when one or both of the variables is of ordinal level
