The Data Analysis Process and Collecting Data Sensibly What is the basis of a study?

The Data Analysis Process and Collecting Data Sensibly

 What is the basis of a study?

How is data classified?

Variable—any characteristic that may change Data—results from observing variables Univariate data—has one variable Bivariate data—has two variables Multivariate data—has two or more variables Categorical data—(qualitative)—has

responses that are types not numbers Numerical data—(quantative)—has responses

that are numbers Discrete data—individual points—integer

values Continuous data—a continuum of data—real

number values

What type a data would the solution set be?discrete or continuous

1. There are 30 spots on a ferry to park vehicles. A car takes one spot and a bus takes 3 spots. What type a data would the solution set be?

2. The fuel efficiency of a new car.

Pg 14 2.1 & 2.2

What is the data analysis process?

Understand the nature of the problem Know the goal and what questions must be answered

Decide What to measure and How to measure it◦ Carefully define the variables and an appropriate means

of determining these values Data Collection

◦ Decide if an existing source will work or if new data should be collected

◦ Determine any limitations of your collection methods Data Summarization

◦ Both graphical and numerical Formal Analysis

◦ Apply appropriate methods Interpretation of Results

◦ Draw conclusions and answer the question

What were the researchers trying to learn?

Was relevant information collected?

Was it collected in a sensible way?

Was it analyzed appropriately?

Are the conclusions supported by the research?

What is the difference between observation and experimentation?

Observational Study—looking at existing characteristics of an existing population and drawing a conclusion

Experimental—if the researcher manipulates one or more factors to determine how a response variable behaves

•Must obtain a sample that represents the corresponding population•Cannot draw cause and effect relationship

•The researcher controls who is in which group and applies a “treatment” on at least one group

Confounding Variable—one that is related to both group membership and the response variable◦ Example

Researchers determined that students who have a larger vocabulary have more cavities.

Is there a confounding variable or is there a causal relationship

age

1. Which of the following statements about experiments are true?

a) a) all experiments must have a control group

b) b) blocking is employed to reduce variation

c) c) Random assignment is only critical for treatment groups as opposed to control groups

d) d) matching can be used in any experiment to eliminate lurking variables

e) e) none of these is true.

1-B

Pg. 20 2.5, 2.6, 2.7, 2.8, 2.9

What considerations need to be made in sampling?

Why would you use a sample over a population?◦Limited resources—time and money◦Destructive measurement—when the

items to be studied requires using up the population Breaking strength of a soda bottle The life of a battery

Random—when all the items have an equal chance of being chosen to be in the sample

Convenience—doesn’t take into account all the factors that may be in the population◦ Just an easy group to gather

Stratified sample—the group is broken into subgroups and then a random sample is chosen from each subgroup◦ By age ◦ By distance

Cluster sample—natural grouping are evident in the population to be studied◦ Need to find out how all university students feel on a topic take a

random sample from each university◦ Zip codes◦ Counties

Systematic—choosing the nth item in the population◦ Every 5th person who walks into the cafeteria

With replacement—the same item can be used over in the sample

Without replacement—the same item may only be used once in the sample

Create a sampling frame—a list of all eligible items in the population identified by a number

Generate a list of random numbers and then use items for your sample

Selection—when a part of the population is excluded from the study because of design

Measurement (numeric) or Response —tends to produce values that differ from the population due to the way something is measured or stated

Non-response—when members of the population who were asked for the opinion do not respond

Start at 8:40

1. A marketing company offers to pay $25 to the first 100 persons who respond to their advertisement and complete a questionnaire regarding displays of their client’s product. This situation is an example of which of the following?a) simple random sample (SRS)

b) convenience sample c) voluntary response sample

d) multistage cluster sample e) None of the above

2. A simple random sample was selected of large urban school districts throughout New England. The selected districts were identified as target districts. Within each district, a SRS of its high schools were chosen and their principals interviewed. Which of the following is not true?a) This is an example of a multi-stage cluster sampleb) Results from the interviews cannot be used to infer responses to the population of interestc) The population of interest is the set of all high school principals from large urban school districts in New Englandd) Not every subset of principals has the same chance of selectione) all are true

1-C 2-B

Pg 26 2.10 and 2.12

What is a comparative experiment?

When the value of some response variable is recorded some method is implemented and the new response is recorded

Experiment—a planned intervention

Variables—factors that can be manipulated

Experimental condition or treatment—that which is done to the variable

Design—the overall plan

Extraneous variables--variables other than the independent variable that may bear any effect on the behavior of the subject being studied but not felt to confound the results

Control—removing extraneous factors

Blocking—creating groups that are similar by removing confounding variables by sharing them among groups

Randomization—ensuring equality of groups

Replication—repetition of an experiment or observation in the same or similar conditions

Control Group—a group that receives no treatment

Blind Study—subjects don’t know which treatment they are getting

Double blind Study—the observer and the subjects do not know who is getting which treatment (only the experimental designer knows)

Placebo—something given that has no effect but doesn’t allow the person to know if they are getting the treatment or the placebo

1. Blocking is utilized to helpa) organize the treatment and control groupsb) counteract the placebo effectc) produce groups that are as similar as possibled) replicate the experiment within each blocke) avoid the need for randomization

2. If the director selects 50 employees at random from throughout the company and categorizes their lunchtime by gender, she is:a) blocking on genderb) testing for a lurking variablec) promoting sexual harassmentd) testing for biase) none of these

1-C 2-B

Pg 35 2.25 and 2.30