Sampling and Experimental Designs

Exam 1

Types of Sampling

• Voluntary response/ self-selected sample

• Random sampling

• Simple random sampling

Voluntary Response/ Self-Selected Sample

• respondents themselve decide whether to be included in the study

• Responses are not representative of the large population

Why are Voluntary response/ self-selected sample not representative of the large population?

These samples are often biased and should not be used to make conclusions about all people/ large populations

The way in which a ______ is chosen has a direct impact on the _____ of the data we collect and what _______.

sample; quality; we are able to do with it

When possible, we want to choose a sample that is ______ of the larger population

Representative

Representative

The characteristics of the sample closely match the characteristics of the larger population

Random Sampling

• The sample are chosen randomly, thus the characteristics of the population are proportionally represented in the sample

Ex. The breakdown of gender in a population is 60-40, then we expect the random samples from the population to also be (approximately) 60-40 regarding gender.

• They are usually computer/random

• Gives the best chance of obtaining a representative sample

Sample error

Any different between the sample result and the population result

Simple random sampling

A simple random sample of size n subjects is selected in such a way that every possible sample of size n has the same chance of being chosen

Types of design studies

Observational studies and Experiments

Observational studies

Researchers observe and measure individuals, but do not modify them in a way

Ex. health care survey, gallop poll

Types of observational studies (3)

• Cross-sectional

• Longitudinal (Prospective) Study

• Retrospective (Case-Control) study

Cross-sectional

data are measured at 1 time period

Longitudinal (Prospective) Study

• data are collected at multiple time points

• Examines trends over time

Retrospective (Case-Control) study

• Researchers go “back in time” and collect data from the past in order to learn more about the current status of a variable/topics

Ex. Records, files; conduct interviews that ask retrospective questions

Ex. What factors contribute to alcohol abuse; what factors are linked to bladder cancer, how old were you when you had your first drink

Experiements

Researchers apply some treatment (modification) to study participants and observe the effects

Clinical trials are well designed experiments that include…

Treatment groups: subjects receive the special/innovative treatment

Control groups: subjects are not given the special treatment, but rather, are treated as usual or are given a placebo

Examples of experiments

• Test the effectiveness of polio

• A treatment group is given a dose of Salk’s vaccine (treatment group)

• A control group is given a placebo treatment that contain s no vaccine

• Researchers compare the incidence rates of both groups, hoping to see fewer vases of disease in the treatment group

What is an important issue in experimental research?

Confounding

Confounding

When the effects of the variables become intertwined so that individual effects of the variables cannot be appropriately identified

How do you minimize confounding

By designing experiments in such a way as to control the effects of extraneous variables

• Blinding

• Blocking

• Randomization

Blinding

Participants don’t know whether they are receiving the treatment or the placebo

• single-blind or double-blind