Assignment 2
Instructions (please read carefully)
General information: This assignment focuses on data wrangling and basic
univariate/bivariate statistics.
Unlike standard assignments, many of the questions will give you multiple possible
approaches – you will need to make decisions, explain your reasoning, and reflect on your
choices.
Submitting the assignment: Please use this word file as your starting point. Add your
answers in the boxes below the questions. Please also copy-paste the R code that you use for
each question. Once you have completed it, convert this word document to pdf and submit
the pdf as well as the R script that you used to come to the answers in Canvas under ->
Assignments -> Assignment 2.
Remember to upload the files to Canvas on Thursday before 12h (noon).
Please name the pdf document and the R script: “A2RIB_TeamName”. For example, if Team
A submitted the files, they would be named A2RIB_TA.pdf & A2RIB_TA.R.
Also note that, unlike the assignment in Week 1, this assignment is graded. Please see the
“Assessment and Grading” section in the Canvas Overview for details.
To check:
1. Set the working directory as your main folder (under Session -> Set Working
Directory).
2. Consult the R instructional videos and the Analysis: “Data-Wrangling A Key Skill”
chapter.
3. Make sure you download the necessary data for this assignment (provided in Canvas).
4. Make sure you download the packages that we introduced in this session, namely
“skimr”, “janitor”, and “kableExtra”.
, Questions
Methodology
1. From the master theses shared on Canvas (module: Master theses examples) identify a
research design that you think involved substantial data preparation or cleaning.
Reflect on the following: has the student described the data cleaning in detail, could
you suggest one way of preparing the data differently and how do you think it could
impact the results?
After skimming over a few theses, we decided to go with “Leadership Explored:
Investigating the Effects of Servant Leadership, Self-Determination Theory, and Perceived
Organizational Support on Organizational Citizenship – A Moderated Mediation Analysis”
with a quantitative study. Another thesis from the Organisation specialisation doing a
quantitative study was considered but ultimately the data preparation and cleaning was less
explicit and detailed, as not as many steps were needed. In the ‘Analytical Strategy’ section
of our chosen paper, data preparation and cleaning were specifically mentioned, as well as
“including missing values and recoding items for consistency” sake. In the ‘Methods’ the
student points out the data preparation of their survey data which included data-cleaning
before the analysis: the initial sample of 225 respondents was reduced to 127 after
excluding participants who did not complete the survey, not consent or lacked employment.
Finally, the sample included 109 participants as these were identified as employees, which
then made up the base for the analysis. The data was prepared in a way that insured
relevance to the study. Further, the Organisational Citizenship Behaviour (OCB) revealed a
low Cronbach alpha value for reliability analysis, which prompted the student to exclude
two items from the scale. The exclusion enhanced the internal validity and increased the
Cronbach value.
One way to prepare the data differently could have been to keep the two excluded items
part of the OCB scale and reported results to compare with the adjusted scale and inform
on the reliability of either scale. Differing results could have been an indication for the
sensitivity of the data and how it is cleaned, while similar results could have strengthened
the conclusion furthermore. Maybe there would not have been a great impact, as Cronbach
alpha went from 0.43 to 0.6 after exclusion, which was described and justified as
‘acceptable’ in this case, but it is still below 0.7 which is typically considered ‘good’.
, 2. From the master theses shared on Canvas, screenshot an example of summary
statistics (e.g., whether in text or table form). Critique the presentation of the
statistics: what works well, what could be improved?
The summary statistics presented here is the table, (above it there was also a brief
descriptive text with the correlation and significance values between some of the
variables). The table, as can be seen in the screenshot, is structured clearly, allowing the
reader to easily assess and skim values as well as significance, especially considering the
different significance of stars – these highlight statistically meaningful correlations. It also
includes the mean and standard deviation for each variable, which is important for the
reader to gain a quick overview and understanding.
One thing making it difficult to read is the number of presented variables and values,
creating a dense summary. Compared to another thesis in the same specialization track and
also doing a quantitative study, the statistical summary is on the shorter side, only
including this table. It could also be helpful to be reminded of the scale range used in the
summary (it is mentioned in the ‘Measures’ section). It might be interesting to get a table of
min and max for data distribution differences for better insight and also for potential
outliers. After the brief text and this table, the student moves on to hypothesis testing, so
the descriptives summary is kept very brief. Regardless, it is generally clear and gives
relevant information.
For this part and for the rest of the assignment, please familiarize yourself with the
billionaires.csv dataset. The data are collected up to 2014. Here are the var descriptions:
Variable name Description
name The name of the billionaire.
rank The rank of this billionaire compared to the rest of the billionaires
reported on. A lower rank means they make more money.
year The year that data about this billionaire was collected.
company.founded The year that the company was founded.
company.name The name of the company.
company.relationship The billionaires relationship to the company.
company.sector The sector of the business, or what segment of the economy they fit
into.
company.type The type of business for this company.
demographics.age The current age of the billionaire. Ages that are represented as -1
stand for ages that were not available in the data that was collected.
demographics.gender A string representing their gender.
Instructions (please read carefully)
General information: This assignment focuses on data wrangling and basic
univariate/bivariate statistics.
Unlike standard assignments, many of the questions will give you multiple possible
approaches – you will need to make decisions, explain your reasoning, and reflect on your
choices.
Submitting the assignment: Please use this word file as your starting point. Add your
answers in the boxes below the questions. Please also copy-paste the R code that you use for
each question. Once you have completed it, convert this word document to pdf and submit
the pdf as well as the R script that you used to come to the answers in Canvas under ->
Assignments -> Assignment 2.
Remember to upload the files to Canvas on Thursday before 12h (noon).
Please name the pdf document and the R script: “A2RIB_TeamName”. For example, if Team
A submitted the files, they would be named A2RIB_TA.pdf & A2RIB_TA.R.
Also note that, unlike the assignment in Week 1, this assignment is graded. Please see the
“Assessment and Grading” section in the Canvas Overview for details.
To check:
1. Set the working directory as your main folder (under Session -> Set Working
Directory).
2. Consult the R instructional videos and the Analysis: “Data-Wrangling A Key Skill”
chapter.
3. Make sure you download the necessary data for this assignment (provided in Canvas).
4. Make sure you download the packages that we introduced in this session, namely
“skimr”, “janitor”, and “kableExtra”.
, Questions
Methodology
1. From the master theses shared on Canvas (module: Master theses examples) identify a
research design that you think involved substantial data preparation or cleaning.
Reflect on the following: has the student described the data cleaning in detail, could
you suggest one way of preparing the data differently and how do you think it could
impact the results?
After skimming over a few theses, we decided to go with “Leadership Explored:
Investigating the Effects of Servant Leadership, Self-Determination Theory, and Perceived
Organizational Support on Organizational Citizenship – A Moderated Mediation Analysis”
with a quantitative study. Another thesis from the Organisation specialisation doing a
quantitative study was considered but ultimately the data preparation and cleaning was less
explicit and detailed, as not as many steps were needed. In the ‘Analytical Strategy’ section
of our chosen paper, data preparation and cleaning were specifically mentioned, as well as
“including missing values and recoding items for consistency” sake. In the ‘Methods’ the
student points out the data preparation of their survey data which included data-cleaning
before the analysis: the initial sample of 225 respondents was reduced to 127 after
excluding participants who did not complete the survey, not consent or lacked employment.
Finally, the sample included 109 participants as these were identified as employees, which
then made up the base for the analysis. The data was prepared in a way that insured
relevance to the study. Further, the Organisational Citizenship Behaviour (OCB) revealed a
low Cronbach alpha value for reliability analysis, which prompted the student to exclude
two items from the scale. The exclusion enhanced the internal validity and increased the
Cronbach value.
One way to prepare the data differently could have been to keep the two excluded items
part of the OCB scale and reported results to compare with the adjusted scale and inform
on the reliability of either scale. Differing results could have been an indication for the
sensitivity of the data and how it is cleaned, while similar results could have strengthened
the conclusion furthermore. Maybe there would not have been a great impact, as Cronbach
alpha went from 0.43 to 0.6 after exclusion, which was described and justified as
‘acceptable’ in this case, but it is still below 0.7 which is typically considered ‘good’.
, 2. From the master theses shared on Canvas, screenshot an example of summary
statistics (e.g., whether in text or table form). Critique the presentation of the
statistics: what works well, what could be improved?
The summary statistics presented here is the table, (above it there was also a brief
descriptive text with the correlation and significance values between some of the
variables). The table, as can be seen in the screenshot, is structured clearly, allowing the
reader to easily assess and skim values as well as significance, especially considering the
different significance of stars – these highlight statistically meaningful correlations. It also
includes the mean and standard deviation for each variable, which is important for the
reader to gain a quick overview and understanding.
One thing making it difficult to read is the number of presented variables and values,
creating a dense summary. Compared to another thesis in the same specialization track and
also doing a quantitative study, the statistical summary is on the shorter side, only
including this table. It could also be helpful to be reminded of the scale range used in the
summary (it is mentioned in the ‘Measures’ section). It might be interesting to get a table of
min and max for data distribution differences for better insight and also for potential
outliers. After the brief text and this table, the student moves on to hypothesis testing, so
the descriptives summary is kept very brief. Regardless, it is generally clear and gives
relevant information.
For this part and for the rest of the assignment, please familiarize yourself with the
billionaires.csv dataset. The data are collected up to 2014. Here are the var descriptions:
Variable name Description
name The name of the billionaire.
rank The rank of this billionaire compared to the rest of the billionaires
reported on. A lower rank means they make more money.
year The year that data about this billionaire was collected.
company.founded The year that the company was founded.
company.name The name of the company.
company.relationship The billionaires relationship to the company.
company.sector The sector of the business, or what segment of the economy they fit
into.
company.type The type of business for this company.
demographics.age The current age of the billionaire. Ages that are represented as -1
stand for ages that were not available in the data that was collected.
demographics.gender A string representing their gender.
