经济代写-ECON 6074
时间:2021-11-07
Midterm Exam
ECON 6074: Causal Inference
Module 2 – 2021/2022


Instructions
• Please submit your work in one zip file containing three files: a) write up of your answers
(including tables), b) log file, and c) do-file that works with your matched data set.
• Please read the manuscript “Demand for Ultrasonography: Roles of Price Framing and Sex
Information” to understand the study setting and design. You will be asked to conduct an
empirical analysis using a sampled dataset from this study.
• Each student is assigned with a unique dataset. Look up your matched Midterm ID number
and download your assigned data set “Midterm_M2_2022_ID[#].dta” from the Moodle
page. If your answers do not match with your assigned dataset (i.e. use another person’s
dataset) you will receive 0 points.
• For description of variables, please refer to “Midterm_M2_2022_codebook.xlsx”.

Questions
1. [20 points] Non-data questions.
a. (5 points) The study involves randomization of two treatments. What are these
treatments? One of the main outcomes is take-up of ultrasonography. How is
this measured?
b. (5 points) What is the identification strategy to find the causal effects of these
treatments mentioned in part a.? What is the key assumption of this strategy?
c. (10 points) There are three hypotheses that are of interest (refer to manuscript).
In part 2, you will be asked to test each of these hypotheses using the provided
dataset. Here, for each hypothesis, write the regression specification that you
will estimate to test the hypothesis.

2. [40 points] Data questions. For each question, report the results in a table and comment
on the findings. Specifically, report the coefficient estimates along with the standard
errors and p-values and comment on the statistical significance. The variables that you
are supposed to use are provided below each question.
a. (10 points) Conduct a balance check with baseline characteristics (listed below)
to test if treatments were randomly assigned across different treatment groups.
Report the results of the balance check in one table.

Baseline characteristics: hhmem1_age educ_year nboys ngirls prefer_son_nxt
prefer_dau_nxt health_good asset_index compute_score heardof_ultrasound
ultrasound_purpose_fetalgrowth ultrasound_purpose_fetalsex
ultrasound_safety_toself ultrasound_safety_tobaby ultrasound_healthinfo_accuracy
ultrasound_sexinfo_accuracy abortion_where married child_die child_lost preg_month

b. (10 points) Test Hypotheses A and B first without fixed effects in the regression
and second with fixed effects in the regression. Do the results depend on the
inclusion of fixed effects? Explain your answer.

Outcome variable: purchase

Fixed effects (enter in Stata command as written here): i.lottery_enu i.svy_center b1.weeks
i.days
c. (10 points) Now always include fixed effects and test Hypotheses A and B first
without control variables in the regression and second with control variables in
the regression. Should we expect the results to depend on inclusion of controls?
Explain your answer.

Outcome variable: purchase

Control variables (enter in Stata command as written here): age_* i.educ_year i.nboys
i.ngirls prefer_son_nxt prefer_dau_nxt health_good asset_index compute_score
heardof_ultrasound ultrasound_purpose_fetalgrowth ultrasound_purpose_fetalsex
ultrasound_safety_toself ultrasound_safety_tobaby ultrasound_healthinfo_accuracy
ultrasound_sexinfo_accuracy abortion_where

Fixed effects: i.lottery_enu i.svy_center b1.weeks i.days

d. (10 points) Test Hypothesis C using the sex preference indicated by the
respondent. Include control variables and fixed effects. Explain your findings.

Outcome variable: wtp

Control variables: age_* i.educ_year i.nboys i.ngirls health_good asset_index
compute_score heardof_ultrasound ultrasound_purpose_fetalgrowth
ultrasound_purpose_fetalsex ultrasound_safety_toself ultrasound_safety_tobaby
ultrasound_healthinfo_accuracy ultrasound_sexinfo_accuracy abortion_where

Fixed effects: i.lottery_enu i.svy_center b1.weeks i.days

Indicator for sex preference: prefer_son_nxt/prefer_dau_nxt







































































































学霸联盟


essay、essay代写