1. Thinking Like An Experimentalist

Question 1

What is an economic experiment?

Answer 1

An observation of human behaviour often with incentivised choice behaviour and often with an experimental environment created and/or manipulated by an experimenter

Question 2

How can we create experiments as an attempt at isolation of a relationship?

Answer 2

If we believe y=f(x1, x2, …, xn) and we wish to explore the relationship between y and x1. We construct an environment in which we aim to observe y, vary the level of x1, control the influence of x2…xn

Question 3

What are the broader purposes of experiments?

Answer 3

Theory testing
Hunting regularities- may become seeds for new theory
Institutional design

Question 4

What did Schelling say about the “matching game” and what year did he say it?

Answer 4

Fully rational players can’t “solve” the game of picking the same answer as their partner but ordinary people seem to be pretty good at it. 86% of people choose heads. 1960

Question 5

Describe the design of Mehta, Starmer & Sugden 1994

Answer 5

-20 similar tasks e.g name a colour
-subjects randomised between two conditions.
-coordinators (motivated), motivated to try to coordinate
-pickers (unmotivated), rewarded as long as they answer all questions

Question 6

What are the possible theories for explanation of coordination success?

Answer 6

Primary salience
Secondary salience

Question 7

What is primary salience?

Answer 7

The tendency for people to choose options with labels that stick out more than others

Question 8

What is secondary salience?

Answer 8

Trying to guess what has primary salience for the other person

Question 9

Results of Mehta, Starmer & Sugden 1994

Answer 9

Higher coordination in motivated coordinators and lower number of response types.
Patterns of results support conclusion that choices aren’t made just from primary salience- there is also secondary salience

Question 10

What should you consider when making an experiment?

Answer 10

Research question?
Hypotheses?
Control variables?
Between or within subjects?
Incentives?

Question 11

What is between subjects?

Answer 11

Subjects are assigned randomly to 2 different groups. Each group experiences only one of the two treatments. We compare responded between the groups

Question 12

What is within subjects?

Answer 12

Every subject experiences all treatments

Question 13

Positives and negatives of within subjects compared to between subjects?

Answer 13

Within subjects generates more data but is less clean

Question 14

How is randomisation used when designing an experiment?

Answer 14

Randomisation can be used to control for a number of possible influences in designs

Question 15

What does Al Roth say is a good experiment?

Answer 15

An experiment that controls for the most plausible alternative hypotheses that might explain what is being observed and therefore allows you to distinguish among them

Question 16

What is internal validity?

Answer 16

A joint property of a specific experimental design and an inference. An inference is internally valid, relative to a given design, to the extent that it is licensed by that design.

The extent to which you can be confident that a cause and effect relationship established in a study can’t be explained by other factors

Question 17

What are some experimental conventions?

Answer 17

-motivate subjects with incentives
-use RIS when subjects do multiple tasks
-create simple environment
-clear instructions
-use written script
-tasks to test understanding
-avoid loaded instructions and deception

Question 18

What did Friedman say in 1953 about experiments in economics?

Answer 18

Unfortunately we seldom test particular predictions in the social sciences by experiments explicitly designed to eliminate what are judged to be the most important disturbing influences. Generally we must rely on evidence cast up by experiments that happen to occur

Question 19

Who was the first person to report results of a controlled experiment in a scientific journal?

Answer 19

Jevans 1870 in nature

Question 20

What do some people attribute the rise in experimental economics to?

Answer 20

Development in information technology allowing these experiments to take place

Question 21

How were lab experiments initially viewed?

Answer 21

Viewed as having very little external validity. Ie results don’t translate to field

Question 22

How did Rubinstein criticise experimental economists on 2001?

Answer 22

He said economists often follow problematic practice since experimental work relies heavily on honesty- Rubinstein 2001

Question 23

What is the goal of an experiment according to “Experimental methods: a primer for economists”

Answer 23

Neither designing an experiment to replicate realism or theory is necessarily right. The goal should be to design an experiment that offers the best opportunity to learn something useful and to answer questions that motivate your research

Question 24

How does “Experimental methods: a primer for economists” view experiments without incentives?

Answer 24

A lab procedure that pays subjects a flat participation fee to respond to hypothetical choices, properly speaking, is a survey and not a controlled experiment because rewards aren’t salient

Question 25

Do rewards change outcomes according to “experimental methods: a primer for economists”

Answer 25

Sometimes salient rewards substantially alter the experimental outcomes and sometimes they don’t

Question 26

What do Jamal & Sunder find about rewards?

Answer 26

The use of rewards tend to increase the reliability of results (Jamal & Sunder 1991)

Question 27

What do Jamal & Sunder find about rewards?

Answer 27

The use of rewards tend to increase the reliability of results (Jamal & Sunder 1991)

Question 28

What are the consequences of holding more variables constant?

Answer 28

Your experiment becomes simpler and cheaper but you learn less about the direct effects and the interactions among variables