Statistics for Premasters, fall 2015 Assignment - Basic terms and exploring data

Q1. For each of the following statements, indicate whether they are true or false, why, and explain with an example.

a) An interval-scale implies that the ratio between values can be meaningfully interpreted.

a. True/false?

b. Why:

c. Example:

b) An ordinal  variable is the same as a nominal variable

a. True/False

b. Why:

c. Example

c) A score of 4 on an interval scale is twice as much as a score of 2.

a. True/false?

b. Why:

c. Example:

d) A confounding variable is a variable that is manipulated by the experimenter.

a. True/false?

b. Why:

c. Example:

e) You cannot meaningfully calculate a mean of a ratio variable

a. True/false?

b. Why:

c. Example: 

Q2. In the following research questions, what is the dependent and what is the independent variable. And what is the measurement level of each variable?

a. Does time spent in front of the TV predict someone's daily calorie intake?

DV: measurement level:

IV: measurement level:

b. At age 4, do boys have more toy cars than girls?

DV: measurement level:

IV: measurement level:

c. Do extroverted people prefer going to a bar to going to the library?

DV: measurement level:

IV: measurement level:

d. Can you use a husband's weight to predict whether his wife is pregnant or not?

DV: measurement level:

IV: measurement level:               

e. Do vegetarians drink more bloody marys (alcoholic drink with tomato juice) than carnivores?

DV: measurement level:

IV: measurement level:               

f. Are people living in Amsterdam, Rotterdam, and Utrecht more beautiful than people living in Nijmegen, Tilburg, or den Bosch?

DV: measurement level:

IV: measurement level:

g. Can the amount of sugar intake predict whether or not someone will ever get a cavity (tooth decay) in his/her life?

DV: measurement level:

IV: measurement level:               

Here's room for a comment in case you feel there is 'reasonable doubt' for one/some of these variables.

Q3. Indicate for each of the following research designs whether the assumption of independence is met. Explain why / why not.

a. In a study on team performance, several different teams that differ in their compositionparticipate. After the team performance task, each team member is asked individually to indicate their subjective feeling of success.

b. In a study on the effect of empathy on performance, participants are asked to make a very difficult test. Half of the participants are allowed to call a friend before making the test and to tell the friend about their feelings about the test, the other half of the participants are not allowed to call anyone.

Q4. Look up at least two different descriptions for 'hypothesis testing' and then write up in your own words what it means (no more than 50 words):

Q5. The following questions concern a dataset on tween mobile phone use (available on Blackboard "dataset_tweens mobile phone 2015"). 1144 children between 9 and 12 answered a survey on their mobile phone use.

a) Construct a histogram of the variable 'Price' (the price that the tweens in this sample paid for their mobile phone).

a. Copy and paste the histogram into the file where you are recording your responses.

b. Explain the terms 'skewness' and 'kurtosis' in your own words

c. Evaluate the histogram in those terms.

b) The following questions concern the number of SMS from friends per week the tweens report to receive (varname = SMS_from_friendsperweek).

a. Calculate the different measures for central tendency (mean, mode, median) and dispersion (min, max and standard deviation).

i. in your opinion, which measure for central tendency reflects the central tendency best and why? Explain on which basis you made your choice (what steps did you follow).

ii. explain in your own words what the standard deviation means, using the standard deviation for SMS_from_friendsperweekas an example.

b. in this dataset, do girls have on average more friends than boys?

c. make a new variable 'from_friend_3_CAT' that consists of a re-categorization of the SMS_from_friendsperweek-variable into 3 categories (few friends, some friends, many friends). Make sure the groups are approximately equal in size.

i. Briefly explain the procedure you used to achieve this categorized variable, what the new categories are, and what percentage of respondents is in each category.

c) The following questions concern the items 'friends_1' to 'friends_11 (11 items in total). These items measure the respondents' feelings of peer pressure. A high score means that you experience a high degree of peer pressure.

a. Make a new variable 'mean_friends' that expresses a mean score of the 11 above items. In order to do so, you first need to check whether you may need to re-categorize some of the items (reverse code them). Explain your procedure.

b. Write down the results for the next few questions in 'scientific' language (as if you were writing a research report):

ii. What is the mean feeling about peer pressure of our respondents?

iii. In this sample do children with a non-European background (variable 'Ethnicity') experience more peer pressure than children from a European background?

iv. What are the confidence intervals for children with a non-European background and for children with a European backgroundfor peer pressure? What can you conclude from the confidence intervals concerning the difference in peer pressure scores between the two groups?

Q5. Graphs

a) Make a bar chart for independent means for your new variable 'mean_friends'that expresses whether there are differences in the mean scores on this variable for girls and boys. Make sure to include error bars and to give the axes proper labels. Paste this figure into your document. What can you conclude from this figure?

b) Make a bar chart for dependent (related) meansfor the variables "sms_to_momperweek" and "sms_to_dadperweek". Paste this figure into your answer sheet. What can you conclude from this figure?

c) Make a grouped scatterplot that shows the relationship between number of friends and price_mobile separately for girls and boys. Also include a regression line for each of the two subgroups. Paste this figure into your answer sheet. What can you conclude from this figure?