ST 311 Sampling Methods and Experiment Terms

Response Variable

Measures an outcome of a study

Explanatory Variable

Explains or influences changes in response variable

Lurking Variables

Not among the explanatory variables but can still influence the intrepretation of the relationship among response and explanatory variable

Confounding Variables

Effects on the response variable cannot be distinguished from each other

Population/Population of Interest

The complete collection of all measurements or data that are being considered

Sample

Subset of members selected from a population

Parameter

A numerical measurement describing some characteristic of a population

Statistic

A numerical measurement describing some characteristic of a sample

Quantitative Data

Consists of numbers representing counts or measurements

Categorical/Qualitative Data

Consists of names or labels

Discrete Quantitative Data

Number of values is finite or “countable”

Continuous Quantitative Data

Infinitely many possible quantitative values, not countable, are also decimals

Convenience Sample

Easy to collect, often have some bias or do not represent the population in general

Simple Random Sample (SRS)

Sample of subjects is selected in a way where every possible sample of the same size has the same probability of being chosen

Sampling Frame

List of units in the population

Stratified Sample

Divides population into two subgroups with the same characteristics and then draws a sample from each subgroup

Cluster Sample

Divides population area into naturally occurring sections, randomly selects some of the sections, then chooses all members from the selections

Systematic Sample

Select a starting point and then select every kth element in the population, works well when units are in a order

Multistage Sample

Collects data by using some combination of basic sampling methods

Bad Sampling Frame

Missing subjects from a list of all the population members

Undercoverage

Sampling frame is missing groups from population/groups have smaller representation in the sample than in the population

Non-response Bias

Some part of the population chose not to respond or were not able to be contacted

Response Bias

Survey responses are not truthful

Wording & Order

Question wording/order may be leading or inflammatory to get a specific response

Completely Randomized Design

Participants are randomly assigned to treatments, is more equal and fair

Randomized Block Designs

Experimenter divides participants into subgroups and then randomly assigns them to treatment groups; variability within is less than variability between groups

Matched Pairs Designs

Experiment only has two treatment groups; participants are paired together based on one or more blocking variables, and then within each pair, they are randomly assigned to different treatments

Mode

Value that occurs with the greatest frequency, only useful for multimodal or qualitative data

Mean

Found by adding all values and dividing it by the number of values in a set, highly affected by outliers, not good for skewed data sets

Median

Value that is in the middle when listed in ascending order, not affected by outliers, can use with any data set

Range

Max data value - min data value, highly affected by outliers

Population Proportion

“p”, a parameter that describes a percentage value associated with a population

Sample Proportion

“p-hat,” a statistic that estimates the population proportion

Mean of the sampling distribution is __ to the population proportion

Equal to