讲解Research Methods、辅导java/python程序设计、讲解HCMC留学生、CS, C/C++讲解

Exam in Research Methods

Individual exam for: Research Methods

Time: December 10th 201 12:00 – December 11th 2018 12:00

This exam is for students with a student number ending in 2, 4, 6, 8, or 10

The exam has five overall questions and additional subquestions. Question 1-3 count 50% towards the final

grade and question 4-5 counts the remaining 50%.

Aides: all aides allowed.

Please ONLY write your UCAS student number on your exam paper

Compile everything in one PDF file. If you want to hand in your R script, you can copy it into the

bottom of the exam document.

The exam is an individual exam. That means that all parts of the exam must be written individually.

Due to the 24-hour format, students are not banned from communicating with each other before handing

in the assignment, but plagiarism of any sort is still considered cheating and will be treated as such.

Good luck!

1

QUESTION 1: (8% weight) (ESTIMATED TIME: 1 HOUR)

In a few paragraphs, please describe your epistemological perspective and discuss how it shapes how you

would conduct research under ideal circumstances.

QUESTION 2: (12% weight) (ESTIMATED TIME 1-2 HOUR)

Below you will find the research methods section of an article about sex work and emotional labor. Please

read it and then provide three points of critique. For example, tell us what is missing, what should be

expanded, what should have been done differently, and/or what you would need to know to evaluate the

reliability, validity and quality of the study. You should limit your answer to 1-2 pages.

“I carried out seven months of fieldwork in three intervals between June 2006 and August 2007 in HCMC

(Ho Chi Min City). During this time, I conducted participant observation in local bars, cafes, sex workers’

homes, malls, restaurants, and on the streets. Scholars who write about working women during the French

and American colonial period refer to women as Hoang Economies of Emotion, Familiarity, Fantasy, and

Desire 259 prostitutes engaged in a form of survival sex in brothels. However, in this article, I focus on sex

workers (Bernstein, 1999) in the contemporary sex industry who choose to enter into sex work rather than

prostitutes or children who are trafficked or forced to work (O’Connell Davidson, 1998). As such, my research

is limited to women over the age of 18 who work as independent agents in local bars and clubs. I began my

research by spending time in local bars and on the streets trying to meet and develop rapport with various

sex workers and clients before asking the women to participate in my project. Once I developed rapport in

the bar scene with the sex workers, I asked if I could spend time with them in their daily lives outside of the

bar. I also relied heavily on the knowledge of Cuong and Loc two motorbike taxi drivers who introduced me

to men and women in the low-end and mid-tier sectors of sex work. Through these interactions, I met a total

of 54 sex workers and 26 clients in the three different sectors who consented to be a part of my project.”

Reference:

Kay Hoang, Kimberly. “Economies of emotion, familiarity, fantasy, and desire: Emotional labor in Ho Chi

Minh City’s sex industry.” Sexualities 13.2 (2010): 255-272.

QUESTION 3: (30% weight) (ESTIMATED TIME: 3-4 HOURS)

Imagine that you are a researcher given the task of doing a research project on the low fertility rates in

Demark, China or both. Please read the case below for more information and feel free to do any additional

research to gain perspective on the topic. In this essay question you need to accomplish five tasks 1) create a

meaningful and useful research question given the case study provided, 2) develop an appropriate qualitative

study design, 3) create at least one tool to operationalize your design, 4) develop a coding strategy and 5)

surmise what kinds of possible findings might your study yield.

You should start by clearly articulating your research question based on the case study provided. Make

certain that the question(s) you ask require a qualitative research study. Do remember that the development

of the research question is very important and shapes the whole study design. You should think carefully

2

about what we already know about this topic and how to go from there. The second step requires you to

develop a qualitative study design. You can choose to do a one-country study or a comparative case study

(between China and Denmark). You must decide what type of research design would be most effective, what

data you must collect to answer the question and what research methods you will use to collect the data.

Most important in this process is to explain why you have chosen to design the study in this particular way

and what kinds of data you expect to collect. You must clearly define your unit of analysis, your study

“population”, your sampling method and desired sample. Third, you must create a tool for operationalization

(either interview protocol, observational guide including description of observation schedule, places and what

is to be observed, a focus group protocol, etc.). Please remember the importance of things like question

design, question placement, choices of observational methods, etc. The tool of operationalization should be

included as an appendix. Fourth, you should discuss how you will analyze the data, what will be your coding

strategy, your coding scheme, what codes might be used and why. Finally, you should think about what

kinds of data your study might yield, and how that data could help inform the government(s) as they work

to develop an appropriate solution to their “fertility crisis.”

THE CASE

One important issue that most developed nations face is a low or declining fertility rate. It has economic,

social and cultural implications. The average birth rate in the European Union is 1.6, well below the 2.1 live

births per woman needed to sustain a population. This is occurring within the context demographic shifts

that include a fast-“greying” population that is retiring and shrinking working-age population. This is a

significant issue because it presents social and cultural challenges and creates potential hurdles for economic

competitiveness.

In Asia, the fertility challenge is also present. This includes Japan, the most economically advanced nation

in the region, which has faced a below-replacement fertility rate since 1973. Similarly, other economically

advanced nations like Singapore and Hong Kong also struggle with this issue. Perhaps more surprising given

its status as a developing country, China has now entered a new phase of below replacement fertility.

Governments have tried a wide-range of campaigns to increase fertility rates from government-sponsored

advertisements, economic and financial incentives, educational programs, and cultural campaigns. However,

despite these efforts, there has been little success. Of course, the problems and solutions for each country

may be different given the different historical path it has taken and varying levels of economic develop. For

instances, in China there are still some negative social and demographic impacts of the one-child policy, most

notably the sex ratio at birth in favour of males. Regardless, governments must consider the context as

they work to develop a mix of financial, economic and/or social programs to shift fertility rates. What kind

of campaign would work in what context? What can we learn from comparing the countries? Specifically

when thinking about China and Demark, both represent types of hybrid systems that coordinate economic

activity, both have a strong state with an important role in organizing social life, and both have a relatively

homogenous population. However, there are significant differences including level of economic development,

cultural practices and historical trajectory. All of this needs to be considered when trying to develop a

successful policy aimed at shifting fertility rates.

3

Question 4 (20% weight. ESTIMATED TIME: 2-3 HOURS)

The quantitative part of the exam focuses on discrimination and attitudes to immigrants in Denmark.

Remember to consider the statistical uncertainty whenever it is relevant as well as interpretations of all

substantial results. Full points will only be given to answers with sufficient interpretations.

In a recently published paper, two researchers designed an experiment to explore the effect of ethnic

discrimination in recruitment in the Danish labor market (Dahl & Krog, 2018). The researchers randomly

assigned traditional Danish or Middle Eastern sounding names to resumes sent for job postings. The resumes

were of comparable quality with only the name of the applicant varying. The design corresponds to the

design used for an experiment presented in the second chapter of the text book.

Question 4.1

Table 1 presents results based on those presented in Table 2 of Dahl & Krog (2018) in the column for men.

The information is presented to align with what you have seen in the course. The outcome of interest is what

proportions of applicants were invited for a job interview

Table 1: Callback rates and N for majority and minority men

Group Proportion called back N

Majority 36.2 207

Minority 19.3 207

Based on the information in the table, what is the estimate of the SATE for men and how do you interpret

it? What assumptions must be met to give this a causal interpretation and are they met?

Question 4.2

Construct and interpret two 95% confidence intervals; one for the majority and one for the minority group.

Based on the confidence intervals, can you conclude if the callback rates for the two groups are statistically

distinct from each other? Why/why not?

Question 4.3

Next, construct and interpret a 95% confidence interval for the difference-in-proportions using the numbers

in the table.

Question 4.4

The article uses applications with randomly assigned names to measure discrimination in the labor market.

Discuss advantages and disadvantages of this approach over other ways to measure discrimination (asking

4

employers if they discriminate, asking employees if they experience discrimination, etc.). Do you think the

results are internally and externally valid?

Question 5 (30% weight. ESTIMATED TIME: 3-4 HOURS)

For this question we will use data for Denmark from the sixth round of the European Social Survey (ESS).

The ESS is a high-quality, multinational survey that is run in a number of European countries every second

year. The dataset includes a subset of the variables in the full dataset. We will use data on age, gender, and

education here as well as information on how respondents view the impact of immigration on the country

and if the respondents see themselves as part of a minority group.

Name Description

Idno Identifier of respondent

Imwbcnt “Immigrants make country worse or better place to live”.

‘0’ = Worse place to live.

‘10’ = Better place to live.

Blgetmg Belong to minority ethnic group in country.

Gender The self-reported gender of the respondent.

Age Age in years

education Education in years

Enclosed with the exam question, you will find a number of data files in csv-format. You must use the data

file with your student number as the name. Each datafile has a sample of 1005 observations, which you must

work with. All students will have different samples and therefore obtain slightly different results. Primoz and

Raphael will be able to help you getting the right dataset.

Question 5.1

Choose appropriate statistics and graphs to describe the variable age.

Question 5.2

Are men or women more skeptical towards the impact of immigration? Make a hypothesis about the difference

for imwbcnt between men and women and use an appropriate test to test it.

Question 5.3

Next make a regression of imwbcnt on gender, age, and education. Interpret the model’s coefficients.

Remember to consider the statistical uncertainty. How much of the variance in attitudes do you explain?

Compare the coefficient on gender with the difference-in-means results in question 5.3.

5

Question 5.4

The variable education measures how many years of education people have. It ranges from 0 to 40 years.

However, there has long been at least seven years of mandatory schooling. In addition, someone who completes

a PhD, the highest educational level, will have had a maximum of 22 years of schooling. Recode education

such that anyone with less than seven years of education is assigned the value seven. Next, recode everyone

with more than 22 years of schooling to the value 22.

Rerun the regression model from question 5.4 with the recoded education variable. Interpret the new model

and compare it with the old model. What if anything has changed?

Question 5.5

The variable blgetmg measures if people see themselves as belonging to a minority group. Create a binary

variable where ‘1’ indicates that someone belongs to a minority and ‘0’ that she does not. Remember to set

those who did not answer or answered ‘do not know’ to missing.

Next, expand the regression model from question 5.5 with the new binary variable that you created.

Interpret the coefficient on the coefficient on the new variable. Are you surprised by this relationship?

Question 5.6

Run a new regression using the same covariates as in 5.6 only this time, you should also include an interaction

between education and blgetmg.

Create a plot where you show the relationship between education and imwbcnt for both those who identify

as belonging to a minority group and those who do not. [Hint: if you predict data, you need to set values

for all variables in the data frame you predict from. You will need to choose values for age and gender in

addition to the values on blgetmg and education.] How do you interpret the results?

Question 5.7

We have seen from the analyses above that age, gender, education, and blgetmg generally predicts attitudes

towards immigration. For each of the variables briefly discuss if there is a causal relationship.

Reference:

Dahl, Malte, and Niels Krog. “Experimental Evidence of Discrimination in the Labour Market: Intersections

between Ethnicity, Gender, and Socio-Economic Status.” European Sociological Review 34.4 (2018): 402-417.