Class Notes (839,150)
Canada (511,218)
POL222H1 (44)
Lecture 6

POL222H1 Lecture 6: Types of Experiments
Premium

4 Pages
71 Views

Department
Political Science
Course Code
POL222H1
Professor
Kenichi Ariga

This preview shows page 1. Sign up to view the full 4 pages of the document.
Description
POL222 Lecture 6 – Experiments Part 2 • Controlled Randomized Experiment o Is a research design which the researcher controls the value of the independent variable o The researcher must manipulate the value ▪ Addresses the concerns of reverse causality and simultaneity o The researcher randomly assigns values of the independent variable to the participants ▪ Addresses the concerns of confounding variable • E.g., (Random Assignment) Participants  Control Group  Y. X=0 • Participants  Treatment group Y. X=1 • By assignment each individual to different group, we would have same people and characteristics in a different group so it’s balanced • After diving them into the group, we will observe the outcome for each group • Even when we randomly assign the participants in a different group, we might invite another in the group unless experiment stimulus in each group o We give a new drug to the treatment group, and a fake drug to the control group o Unless we do this, we cannot really find the causal impact in the new drug • In order to single out the impact of negative message, we have to show something similar to the control group that might have a neutral or positive message • Without new content we cannot single out a single impact for the bias • We have a complete control over that makes a clear order over X and Y • All the confounding variables observed or not have to be balanced between treatment or controlled • Experimental stimulus  should be careful how we control experimental stimulation • Different Types of Experiments • Lab Experiments o Conducted in a lab setting o Recruit voluntary participants • Field Experiments o Conducted in a real-world setting o Become more popular today • Survey Experiments o A sample of people randomly selected from a large population ▪ E.g., population  random sample  through [random assignment] to control group or  treatment group o Most basic is population of people in sample are equal o We can learn the characteristics of relationship in population and in the sample o Random sample will provide you representative sample  tell you the relationship in the population o Random sample helps you choose from population o Random assignment is a procedure which you will divide when you try to identify causal relationship • Survey Experiments for Causal Inference • Question: Does conscription reduce the public support for war? o Contradictory Theoretical Arguments: ▪ Conscription would reduce support for war, as larger segment of population must bear a direct cost of the war and perceive greater risk • Everyone may have someone fighting for war for themselves so they would be careful about the war • Many would oppose going to war ▪ Opposite: Conscription would increase support for war, as larger segment of population has a direct stake in the outcome, and causalities are born fairly • Importance for war and this increase support for war because many people would consider that the country is engaging • Conscription  Public Support for War • Nature of War (may also be a confounding variable)  Conscription • They needed to involve a large segment of the population thus it would affect how families support the war or if people are actually engaged • In an idea scenario, if you can randomly assign rather the country can use conscription or public support and how each country will feel about it …but that’s impossible o Because we cannot randomly assign to each country • A survey of approx. 2000 randomly-sampled individuals in the US o Randomly manipulate beliefs about the likelihood of conscription in near future ▪ Treatment: tell respondents a news story about members of Congress calling for reintroduction of military draft
More Less
Unlock Document

Only page 1 are available for preview. Some parts have been intentionally blurred.

Unlock Document
You're Reading a Preview

Unlock to view full version

Unlock Document

Log In


OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


OR

By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.


Submit