Week 10 Reading Assignment

Week 10 Reading Assignment
Read the textbook (Chap 8: Experiment for Evaluation Research), read the notes on experiment (placed in the Week 10 Reading Assignment [2] folder), watch the video (placed in the Week 10 Reading Assignment [2] folder) and answer the following questions. https://www.youtube.com/watch?v=D3ZB2RTylR4&feature=emb_title
Week 10: Experiment – Notes

Chap 8: Experiments for Evaluation Research

1. Experiment

A method for studying (or testing) hypothesized causal relationships (i.e., hypotheses) under carefully controlled conditions.

2. Characteristics

The most powerful method to study cause-and-effect relationship.
Permits a great deal of control over subjects and the setting.
The most artificial method of data collection.

3. Procedure

The researcher systematically manipulates the independent variable (treatment).
S/he observes if a change in the independent variable (treatment) causes a change in the dependent variable.
S/he tries to exert control over the conditions of the experiment and to eliminate the possibility of extraneous variables influencing the dependent variable.

4. Classical Experimental Design

Subjects

/ \

Randomly assign

/ \

Experimental Group Control Group

| |

Measure Measure <= Pretest

Depend Var Depend Var

| |

Treatment No treatment <= Independent variable

| or “placebo”

| |

Measure Measure <= Posttest

Depend Var Depend Var

5. Matching

Random assignment <====> Matching

| |

Random sampling <====> Stratified sampling

5. Double-blind test

Neither experimenters nor subjects know which, placebo or actual treatment, the subjects are receiving.
Often used in medical experiments.

6. Assessment of Experiment

Internal validity
The ability to determine that no other variable but the independent variable (treatment) actually produced the effects on the dependent variable.
3 ways to increase internal validity
Use of a control group
Control of the experiment
Treat control and experimental groups in the exactly same way except the experimental treatment (i.e., the independent variable)
Randomization

External validity
The ability to generalize findings from a specific setting & sample to a broader range of settings and people.
Two ways to increase external validity
Sampling
Use a probability sampling design, if possible, to select subjects.
Replication
Repeat the experiment in different settings or using different samples of subjects.

7. Non-lab (Field) Experiment

An experiment that takes place in the “field” – in natural settings.
Example
Experimental evaluation research
Weakness
Low control and low internal validity
Ethical problems

w10r1
What is the purpose of classical experiments?

w10r2
Why in classical experiments must everything other than the dependent variable be tightly controlled?

w10r3
In classical experiments, how does a control group differ from a treatment (i.e., experimental) group?

w10r4
What is evaluation research?

w10r5
In evaluation research, what is the independent variable?

w10r6
In classical experiments, why do you have to treat each participant in exactly the same way?

w10r7
Suppose a company wanted to evaluate the effect of a longer lunch break on employee productivity. You were asked to conduct evaluation research using a pre-experimental design. In this study, you found that lengthening the lunch break increased employee productivity. That is, in comparison with the employee productivity in the period in which workers took a one-hour lunch break, the productivity increased dramatically in the period in which workers took a two-hour lunch break. Given this result, can you say with confidence that the longer lunch break caused the increase in worker productivity? Explain your answer.

w10r8
Give one example of ethical problem a researcher may encounter in a classical experiment or experimental evaluation research.