Slides to accompany Weathington, Cunningham & Pittenger (2010), Chapter 3: The Foundations of Research 1.

Slides to accompany Weathington, Cunningham & Pittenger (2010),

Chapter 3: The Foundations of Research

1

Objectives

• Hypotheses and research

• Utility of hypotheses

• Types of hypotheses

• Measurement

• Reliability of measurement

• Validity of measurement

• Populations and samples2

What is the difference between a theory and a hypothesis?

3

Hypotheses

• Are informed, specific predictions about how multiple variables will be related

• Based on theory or previous research

• Guide the progress of science

– What data to collect

– What research methods to use

– How to analyze the data4

A theory is…

• A broad set of general statements and/or claims that helps us to explain and predict events

• Not the same thing as a hypothesis

• May develop from a series of studies that test hypotheses

5

Role of Theory, Hypothesis, Research

THEORY

RESEARCH

HYPOTHESIS

PREVIOUS RESEARCH

Support or refute theory

Basis of additional research

6

Confirmatory Research

• When goal is to support or confirm the validity of an existing theory

• Hypotheses are used in these types of studies

– Helps to protect against the Idols of the Theatre

7

Exploratory Research

• Focus is on examining an interesting phenomenon

• Prior theory is not required

• Caution against Idols of the Cave

– Systemmatic observation can and should still be used

8

Utility of Hypotheses

• Guide to specific variables

– Dependent (DV) vs. independent (IV)

– Subject

– Control

• Describe the variables’ relationship(s)

– Causal or correlational?

• Link research to population

9

Types of Hypotheses - #1

Estimating population characteristics

• Inferring population details from sample

1. Data collected from sample

2. Descriptive statistics calculated

3. Infer to the population level

• If sample is truly representative

• Statistics are always estimates of parameters

10

Types of Hypotheses - #2

Correlational

• X and Y are related

• Positive vs. negative relationship

• Strength of the relationship

11

12

Types of Hypotheses - #3

Difference among Populations

• Testing for differences between average members of separate populations

• 1 variable classifies members of groups, another variable is the DV of interest

• Sample statistics to make inferences about population-level differences

13

Types of Hypotheses - #4

Cause and effect

• X Y

• Causal relationship supported if:

1. X before Y in time

2. X, Y are correlated

3. No other variables explain X Y

14

Measurement is...

• a way of quantifying our observations

• objective

• replicable

15

Operational Definitions

• Formula for a construct that other scientists can use to duplicate it in future studies

– Focus on observable signs of constructs

– Not simple description

1. From hypothesis, identify the constructs

2. Choose a form(s) of measurement that allows us to address each construct best

16

Anxiety

OBSERVABLE EVENTS

HYPOTHESIZED CONSTRUCT

No opportunity for escape

Unfamiliar situation/change

Threat of aversive eventPhysiological measures:•Sweating•Increased heart rate•Increased blood pressure•Rapid, shallow breathing

Behavioral measures:•Inhibited behavior (unable to perform simple tasks)•Pacing, rapid eye movements, startle-reflex•Stammering, other speech problems

Other indicators:•Responses to “anxiety assessment”•Expressed need to be with others

17

Measurement Scales

• Nominal = qualitative, categories

– Measures the property of difference

– Sorts objects/attributes into categories

– Please indicate your sex: M F

• Ordinal = quantitative, ranks

– Measure differences in magnitude

– Grade scale A > B > C, but how much?

18

Measurement Scales

• Interval, quantitative

– Different, magnitude, and equal interval

– Can add and subtract

– Personality test

• Ratio, quantitative

– Diff., magnit., equal ints., true 0

– Time on task19

Reliability of Measurement

• Maximum consistency is the goal

• Challenges:

– Measurement error - random

– Bias error – consistent/constant

Score = True score +/- Measurement error

• Table 3.520

Validity of Measurement• How accurate is the measurement?

– “Trueness” of the interpretations researchers make from the test scores

• Judgment call based on data

– Face validity = authenticity?

– Content validity = true behavior sampling?

– Predictive/concurrent validity = X, Y relationship as expected?

– Construct validity = accurate construct measurement?

21

Unreliable and Invalid

22

Reliable, but Invalid

Measure can be reliable, but still be invalid

23

Reliable and Valid

Measure must be reliable to be valid

24

Relating Samples to Populations• Samples = smaller set of the larger

population of interest

– Representative vs. convenience

•Size and matching characteristics

– Manageability

•Resources and costs

25

Population SampleSampling frame

26

Random Sampling

• Best way to generate a representative subset of a population

• Simple random sampling = Each member with an equal probability of being sampled

27

Random Sampling

• Essential when:

1. Goal is to estimate pop. characteristics

2. Trying to develop a test or intervention for a larger population

28

Random Sampling

• Not essential when we are interested in basic relationships among variables

• BUT,

– Risky generalization

– Requires multiple replications

29

Random Sampling from Population

POPULATION

SAMPLE

INFERENCE

30

Random Assignment

• Not same as random sampling

• Assignment = actual placement in experimental groups

• Minimizes confounds and maximizes transfer of results to pop.

– Good for internal and external validity

31

No Confounding Variables

SAMPLE

Control Group

Experimental Group

Differences are due to manipulation, not an extraneous variable

because mood is randomly determined.

32

Confounding Variables

SAMPLE

Control Group

Experimental Group

Unclear if differences are due to manipulation or confounding variable

(mood)33

Random Assignment

SAMPLEControl Group

Experimental Group

Now you can test these two groups for differences with less

concern for confounds 34

Questions so far this chapter?

What about so far in this course?

35

What is Next?

• **instructor to provide details

36