# STA 261 Study Guide - Final Guide: Standard Deviation, Contingency Table, Design Patterns

291 views11 pages

Published on 4 Jul 2016

School

Department

Course

Professor

Module 1:

Descriptive vs. Inferential Statistics

oDescriptive- summarizing the data, Mean, Median, Standard Deviation,

Graphical Displays

oInferential- drawing conclusions about a population based on a sample,

confidence intervals, hypothesis tests

Statistical Question: a question that can be answered by collecting data- can’t have

an answer that is just based on a single number.

Sample vs. Population

oSample- a subset, or part of the entire group of individuals of interest

oPopulation- entire group of individuals or items of interest

Parameter vs. statistic

oParameter- a value that is computed from the whole population (a fixed

value)

oStatistic- a value that is computed from the sample (a variable value- changes

from sample to sample)

Module 2:

Categorical vs. Quantitative Data

oCategorical:

Can be numerical

Doesn’t make sense to add, subtract, or average the values

oQuantitative:

Are numerical

Adding, subtracting, or averaging the values has meaning

Possible to change into qualitative data when looking at groups

Two different types: discrete and continuous

Discrete- finite or countable (# rooms), isolated points on a

number line, associated with counting

Continuous- an interval or a collection of intervals of real

numbers (height), any number in a continuous interval, the set of

all possible values is in an interval of numbers, associated with

measuring.

Data Displays for Catergorical Data:

oFrequency and relative frequency tables

oPie Charts

oBar charts and pareto charts

Displays for Quantitative Data:

oDotplots

oStemplots

oHistograms

oTime Series Plots

find more resources at oneclass.com

find more resources at oneclass.com

oBoxplots

Measures of Spread

oRange= Maximum value – minimum value

Reported when reporting mean, median, or mode

oIQR= Q-Q1

Reported when reporting median

oStandard Deviation

Reported when reporting means

SOCS

oShape: mention modality and skewness

oOutliers: use fences to help you determine outliers

Lower Fence= Q1-1.5(IQR)

Upper Fence= Q3+1.5(IQR)

oCenter: report which measure of center that best describes the “typical” value

Mean, median, or mode

oSpread: report the measure of spread that is paried with your meaure of center

Mean=Standard Deviation

Median=IQR

Mode=Range

Z-Scores

oA measure associated with each observation that indicates the distance from

the population mean in standard deviations

oA measure of relative standing

oCan be used to compare observations from different populations

Z= (observation-mean)/(standard deviation)

The Empirical Rule

oOnly applies to approx. normal distributions

oApprox. 68% of all obs. w/in 1 std. dev. Of mean

oApprox 95% of all obs. w/in 2 std. dev of mean

oApprox. 99.7% of all obs. w/in 3 std. dev of mean (Almost all obvesvations

Module 3:

Obsevationa Studies can only reveal association or correlation

Retrospective Study- conducting an analysis on the past

oIdentify subject with a specific characteristic and then look into their history

to find thins that may be related

oProne to errors

oExample: Case- control study (used in medical research)

Prospective Study- identifying subjects in advance and then collecting data as

events unfold

oCohort Study

Sample Survey- takes cross section of a population at a given time

find more resources at oneclass.com

find more resources at oneclass.com

oCross-Sectional Study

Designed Experiments- researches impose treatments and controls

oHelp to establish causation

oCalled randomized, comparative experiments

Factor- a variable of interest

Levels of a factor- the specific values the experimenter chooses for a factor

Treatment- the combination of specific levels from all the factors that an

experimental unit receives

Control Group- the group that participates in the experiment and whose treatment is

nothing

oUsed to compare results to other levels of the factors

Blinding- disguising the prupose of the experiment, disguising the different levels

of an experiment, etc. Two main classes of individuals who can affect the outcome

of an experiment. Those who could nfluence the results, those who evaluate the

results.

oSingle-blind experiment

When all the individuals in either one of the classes mentioned above

are blinded

oDouble-blind experiment

When everyone in both classes are blinded

Placebo- a “fake” treatment that looks just like the treatments being tested

The Placebo Effect- subjects treated with a placebo sometimes improve

Blocking and Matching

oUsed when groups of experimental units are similar

oIsolates the variability due to the differences between the block so that we

can see the differences caused by the treatments more clearly

oRandomized Block Design- randomization occurs only within blocks

oMatching is often used in observational studies

Subjects are paired because they are similar in ways that are not being

studied

Matching reduces variation, similar to blocking

4 Principles of Experimental Design

oControl

Control sources of variation by making conditions as similar as

possible for all treatment grouos

oRandomization

Allows us to equalize the effects of unknown or uncontrollable sources

of variation

oReplication

Apply each treatment ot a number of subjects

find more resources at oneclass.com

find more resources at oneclass.com

## Document Summary

Descriptive vs. inferential statistics: descriptive- summarizing the data, mean, median, standard deviation, Graphical displays: inferential- drawing conclusions about a population based on a sample, confidence intervals, hypothesis tests. Statistical question: a question that can be answered by collecting data- can"t have an answer that is just based on a single number. Sample vs. population: sample- a subset, or part of the entire group of individuals of interest, population- entire group of individuals or items of interest. Parameter vs. statistic: parameter- a value that is computed from the whole population (a fixed value, statistic- a value that is computed from the sample (a variable value- changes from sample to sample) Doesn"t make sense to add, subtract, or average the values: quantitative: Adding, subtracting, or averaging the values has meaning. Possible to change into qualitative data when looking at groups. Discrete- finite or countable (# rooms), isolated points on a number line, associated with counting.