Unit 3 The Theme of Statistics

Population

Population

The entire group you want to study

Census

A study of the entire population of interests

Parameter

A value calculated for the entire population (usually unknown)

Sample

Subgroup of the population we actually study

Statistic

A value calculated from the sample data only (always known)

Bias

For different reasons you consistently over or underestimate (our center is off)

SRS (simple random sample)

Each person is equally likely to be selected (RNG, D-table)

Stratified Random Sample

When we want to account for characteristics (Split population into stratus then do an srs within each state, selecting some members from all of the groups)

Cluster Sample

When we have a very large population. Its convenient grouping. You randomly select entire groups and include all units of each group in your sample (put population into groups —> clusters, then do srs on the clusters)

Example: Rows/columns

Systemic Random Sample

When we cannot do SRS or cluster (randomly pick first then every 10th or 50th goes in your sample).

Undercoverage Bias

Occurs when a part of the population is excluded from your sample

Non-response bias

Occurs when those who opt out of a survey are systematically different from those who complete it, in ways that are significant for the research study

Response bias

A general term used to describe a number of different conditions or factors that cue respondents to provide inaccurate or false answers during surveys or interviews

Voluntary response bias

Occur when your sample is made of people who have volunteered to take part in the survey (more likely to have an opinionated sample)

Convenience Sample bias

A non-probability sampling method where units are selected for inclusion in the sample because they are the easiest for the researcher to access

Observation study

research studies in which researchers collect information from participants or look at data that was already collected (no assignment at treatment, subjects make decisions for themselves, never claim cause and affect)

Experiment

An ordered procedure which is performed with the objective of verifying, and determining the validity of the hypothesis (assignment of treatments, tell subjects what to do, may claim cause and affect)

Response variable

What we measure

Explanatory Variable

Changes in the explanatory variable may help explain changes in the response (what you manipulate or observe changes in).

Factors

The variables in the study that we believe will influence the results

Levels

A classification that relates the values that are assigned to variables with each other

Treatments

The application of statistical methods and techniques to analyze and interpret data in various fields

Randomization

Random assignment of treatments (cause and affect)

Equalization

Control over all other variables (fix confounding)

Replicate

Get largest sample size possible (reduce variability/more precise)

Comparison

Must be able to compare

Confounding

Occurs when 2 things change together, and we cannot tell which factor is responsible for the change in response.

Blind

One party doesn’t know which treatment is what (subjects or administration)

Double blind

Two parties don’t know which treatment is what

Placebo

Something that appears to the participants to be an active treatment, but does not actually contain the active treatment.

Control Group

The group in an experiment that does not receive any change in the variable. This group is left as natural as possible and used as a control to see if there is a change from the normal results.

CRD (completely randomized design)

One where the treatments are assigned completely at random so that each experimental unit has the same chance of receiving any one treatment (assigning treatments, when subjects cant/wont do both treatments, we don’t know relevant information about the subjects).

Crossover

A repeated measurements design such that each experimental unit (patient) receives different treatments during the different time periods (do treatments over time).

Paired Design

A type of randomized block design that has two treatment conditions and pairs subjects based on common variables, such as age, grades, health level, or sex (subjects can do both treatments are the same time).

Block design

The arrangement of experimental units or subjects into groups called blocks. This design is typically used to account for or control potential sources of undesired variation (know something that affects the response).

Blocks

Similar for that known factor

Random Sample

Inference about the population

Random Assignment

Inference about cause and effect

Population of Interest

The population from which the researcher wants to draw conclusions.

Outlier

A single data point that goes far outside the average value of a group of statistics.

Statistically significant

Greater than what might be expected to happen by chance alone.

Sampling variability

The extent to which the value of a statistic differs across a series of samples, such that there is some degree of uncertainty involved in making inferences to the larger population.