Statistical Methods I - Chapter 1

Experimental Design Issues

Lack of detail, lack of definition of what you're testing, group too large and diverse.

What are 2 types of data

Qualitative and Quantitative

Simple Random Sample

-Random numbers generated by a random number table, a software program or calculator. -Assign a number to each member of population -Members of population with a matching number become part of the sample.

What is Data

Information coming from observations, counts, measurements or responses

Inferential Statistics

Involves using sample data to draw a conclusion. In a study 70% of unmarried men were alive at 65 while 90% of married men were alive at 65. You can infer that married men are healthier or married men live longer.

Descriptive Statisitcs

Invovolves organizing, summarizing & displaying data. For example: In a study 70% of unmarried men were alive at 65 while 90% of married men were alive at 65.

Double-blind

Neither the subject nor the experimenter knows if the subject is receiving a treatment or placebo.

Quantitative Data

Numerical measurements or counts. For example: age, weight of a letter, temperature

What are 4 data collection methods?

Observational Study, Experiment, Simulation, Survey

Confounding Variables

Occurs when an experimenter cannot tell the differences between the effects of different factors on a variable. (2 things are happening at once such a group takes a new diet pill and also has a strict diet. you wouldn't know which was making the participant lose weight...the pill or diet.)

What are two types of data sets

Population and Sample

Randomization

The process of randomly assigning subjects to different treatment groups

Statistics

The science of collecting, organizing, analyzing and interpreting data in order to make a decision

Simulation

Uses a mathematical or physical model to reproduce the conditions of a situation or process. Often involves computers.

Replication

repetition of an experiment using a large group of subjects. Do on a small scale then on a larger scale. Good example is testing vaccines on 10,000 people then again on another 10,000.

Cluster Sample

Divide the population into groups (clusters) and select all of the members in one or more, but not all, of the clusters.

Parameter

A numerical description of a population characteristic (Population ->Parameter)

Statistic

A numerical description of a sample characteristic (Sample -> Statistic)

Observational Study

A researcher observes and measures characteristics of interest of part of a population.

Placebo effect

A subject reacts favorably to a placebo when in fact he or she has been given no medical treatment.

Sample

A subset or part of the population. For example: All female adults, the 1242 undergrads in college and live outside the us

Blinding

A technique where the subject does not know whether he or she is receiving a treatment or placebo

Experiment

A treatment is applied to part of a population and responses are observed

Survey

An investigation of one or more characteristics of a population. Commonly done by interview, internet, phone or mail.

Random Sample

Every member of the population has an equal chance of being selected.

Sample Size

The number of subjects in a study. The bigger the sample size the better.

Systematic Sample

Choose a starting value at random. Then choose ever kth member of the population. For example: Assign a number to each household in a town, choose a random starting number. Then select every 100th household from that number.

Convenience Sample

Choose only members of a population that are easy to get to. For example: students leaving the library.

Qualitative Data

Consists of attributes, labels or nonnumerical entries. For example: major, place of birth, eye color

Control

Control effects being measured

4 Key Elements of Experimental Design

Control, Randomization, Sample Size, Replication

What are the 2 branches of statistics

Descriptive and Inferential

Stratified Sample

Divide a population into groups (strata) and select a random sample from each group.

Randomized block design

Divide subjects with similar characteristics into blocks and then within each block randomly assign treatment groups. For example divide the group by age group then make selections.

Complete randomized design

Subjects are assigned to different treatment groups through random selection

Matched-Pairs Design

Subjects are paired up according to a similarity. One subject pair is given treatment the other is not. Happens with twins a lot.

Population

The collection of all outcomes, responses, measurements or counts that are of interest. For example: All adults in the US, All undergraduates in college and live outside the us.

6 Components of Designing a Statistical Study

identify variables, develop a detailed plan for collecting data, collect the data, describe the data using descriptive statistic techniques, interpret the data and make decisions about it, identify any possible errors