Ch 5: Measurement Concepts Ch 6: Observing Behavior.

Ch 5: Measurement Concepts Ch 6: Observing Behavior

Reliability Reliability refers to the consistency or stability of a

measure of behavior [p92]

If you weighed yourself now and then at the end of class and you weighed the same both times, you would say the scale is reliable.

True score: The real score on the variable

Measurement error: occurs when a measure yields inconsistent results; the greater the inconsistency, the greater the measurement error

Reliability of Measures

How can we assess reliability?

• Correlation coefficients tell us how strongly two variables are related.

- Pearson Product-Moment Correlation Coefficient (noted as r in text) [p93]

Coefficients range from 0.00 to - 1.00 and 0.00 to +1.00

Sign of the coefficient indicates direction

Value of the coefficient indicates the strength

Reliability of Measures

- 1.00 + 1.000.00

Variables covary in opposite directions

Variables covary in the same direction

Methods of assessing reliability [pp 94-96]

Test-retest reliability: Assesses reliability of a score by measuring the same individuals at 2 points in time

Internal consistency reliability: Assesses reliability of the assessment tool (e.g. test) at one point in time Questions should yield consistent results Split-half reliability – individual’s total score on one half

of the test is correlated with the total score on the other half of the test

Methods of assessing reliability: Interobserver (Interrater) Reliability

A measure of how often two or more observers agree (are consistent) in their observations. [p96]

Nominal scale: percentage agreement

How we assess interobserver reliability:

Number of times two observers agree

____________________ X 100 Number of opportunities to agree

Construct Validity of Measures [pp 97-100]

Refers to the adequacy of the operational definition of variables

Is the measure that is used actually assessing what it is supposed to assess?

If so, it has face validity.

Example: Are facial expressions an adequate measure of happiness?

Indicators of construct validity

Face Validity: The content of the measure appears to reflect the construct being measured.

Content Validity: The content of the measure is linked to the universe of content that defines the construct• Facial expressions are part of a set of behaviors related

to happiness, such as body posture, thoughts, etc.)

Example: Are facial expressions an adequate measure of happiness?


Predictive Validity: Scores on the measure predict behavior on a criterion measured at a time in the future• SAT (Scholastic Aptitude Test predicts future success at

a university) Concurrent Validity: Scores on the measure are

related to a criterion measured at the same time• People who score high on a verbal anxiety test

experience increased sweating at the same time

• People who have happy facial expressions concurrently report feeling happy


Convergent Validity: Scores on the measure are related to other measures of the same construct.• A score for happy facial expression is related to a

score for body posture or mood or optimism Discriminant Validity: Scores on the measure

are not related to other measures that are theoretically different.• A score for happy facial expression is not related to

one for intelligence or cleanliness

Measurement Scales

Four levels for quantifying behavior: nominal, ordinal, interval, and ratio [p102-105]

Nominal scale

Classifies behaviors, events, and characteristics into different categories [p102]

No numerical or quantitative properties

Independent variables are often nominal or a categorical variable

OrdinalScale

Measurement scale in which events and behaviors can be rank ordered (i.e, first, second, third, etc.) [p103]

• Allows categories to be ordered first to last, highest to lowest, biggest to smallest, etc.

• Quantitative but no values attached to the intervals

Interval Scale

Measurement scale that allows researcher to specify how far apart two observations are on a given dimension [p104]

Difference between the numbers is meaningful• Intervals are equal in size

• Quantitative but no meaningful zero reference point

Ratio Scale

Measurement scale that is quantitative, with all numerical properties including an absolute zero reference point [p104]

Let’s practice! Identify the measurement scale for the following data:

Circle your marital status: Married Single Divorced Engaged


Circle your marital status: Married Single Divorced Engaged

NOMINAL


Do you go to work? Yes No


Do you go to work? Yes No

NOMINAL


If you work, how many hours a week do you work? _______


If you work, how many hours a week do you work? _______

RATIO


Rate your enjoyment of college on the scale below.

1 2 3 4 5

Not Very Much Very Much


Rate your enjoyment of college on the scale below.

1 2 3 4 5

Not Very Much Very Much

INTERVAL


What is your class standing?

Freshman Sophomore Junior Senior


What is your class standing?

Freshman Sophomore Junior Senior

ORDINAL

Places to Observe on Campus (Spring 2010) 1. Area in the middle of campus, by the Info Trolley. 2. By the food place on campus which is near the gym/pools. 3. By the statue of the boy overlooking the water feature. 4. Entrance of the SE parking structure (people exiting structure) 5. Entrance of the SE parking structure (people entering structure) 6. Entrance of the SW parking structure (people exiting structure) 7. Entrance of the SW parking structure (people entering structure) 8. By the entrance of the bookstore. 9 In front of C Building, by mirror pools 10. Shuttle stop by U Building 11. Bus stop on Colorado in front of library 12. Passenger drop off/pick up (Colorado B. in front of L Bldg) 13. Starbucks

Reactivity

Occurs when individuals change their usual behavior, when they know they are being observed [p101]

Minimize reactivity by:• Allowing time for individuals to become

used to the presence of an observer or the recording equipment

Time sampling: Researchers choose time intervals for making observations

• Systematic: Schedule observations to occur at a regular time

• Random: Use some random means for identifying times for observations

• Event sampling is used for rare events

Situation sampling: Researchers choose different settings, circumstances, and conditions for their observations

Sampling Behavior [p116]

Classification of Observational Methods

Two categories of observational methods:

• Observation without Intervention

• Observation with Intervention

Observation without Intervention

Naturalistic Observation (AKA: field work or field observations): Observation in natural (“real-world”) settings without an attempt to intervene or change the situation [p110]

Participant observation

Observer is an active participant in the natural setting he or she observes [p112-113]

Norah Vincent

“My Life as a Man”

• Undisguised (unconcealed): people in the setting know they are being observed

• Disguised (concealed): people don’t know they are being observed

Observation withoutIntervention

Physical traces: The remnants, fragments, and products of past behavior

Products: Creations, constructions, or other artifacts of earlier behavior

Natural-use traces

Produced without any intervention by the investigator

Controlled-use traces

Produced with some degree of intervention or manipulation by the investigator

Natural use or controlled use trace?



Archival records

Public and private documents that describe the activities of individuals, institutions, governments, and other groups [p204]

Archival Research [118-121]A non-reactive measure (or indirect method) for collecting data—when the individual who did the behavior is no longer present

Archival research involves using previously compiled information to answer research questions

Running records & records of specific events

Running records are continuously kept and updated (e.g., check book)

Records of specific events (e.g., diploma)

Archival data are used to:

test hypotheses as part of a multimethod approach

test the external validity of laboratory findings

test hypotheses about past behavior assess the effect of a natural treatment

Multimethod approach

Researchers use a variety of measures to examine a research question

Natural treatments

Naturally occurring events that impact society and individuals

Selective deposit

Occurs when some information is selected to be included in the archival record, but other information is not

Selective survival

Occurs when information is lost or missing from an archival source

Content analysis

The process of making inferences based on objective coding of archival data [p120]

Quantitative analysis

Classifying events and behaviors into categories to count their frequency of occurrence [p109]• Assign numerical values to responses and measures and

then subject the data to quantitative statistical analyses

Ex: Count the number of times gender-stereotypical jobs were assigned to characters in the story.

Qualitative analysis Subjective judgments about the content in an

archival record [p109]• Describe behavior or findings based

on themes that emerge from the data.

• Data are nonnumerical and expressed

in language and images

Ex: Watch the tape of Osama Bin Laden and tell me if you think he is being deceptive.

Three Steps of Content Analysis

1) Identify a relevant source

2) Sample selections from the source

3) Code units of analysis

Observation with Intervention [p114]

Systematic observation: The careful observation of one or more behaviors in a particular setting. Use when: • interest is in only a few very specific

behaviors

• observations are quantifiable

• researcher has developed prior hypothesis

Field experiment

Researcher manipulates an independent variable in a natural setting and observes behavior (dependent variable) [pp 83, 113]

Coding systems for systematic observation… [p115]

are either (1) developed to fit the needs of the particular study or (2) “borrowed” systems developed by others

should be as simple as possible

must allow researchers to easily categorize behaviors

Coding systems can involve:

Comprehensive records of people’s behavior (e.g., complete records, such as video tapes)—a qualitative record

Selecting specific behaviors to record—a quantitative record [pp115-116]

Equipment & Narrative RecordsUsed when researchers want a complete (comprehensive) reproduction of people’s behavior [pp 115-116]

• Examples: video and audio recordings, field notes

Systematic Observation: Methodological Issues [pp115-116]

1. Equipment: can fail

2. Reactivity: the probability that the presence of the observer will affect behavior

3. Reliability: refers to how stable/consistent the measure is over time or between observers

4. Sampling: refers to how participants and behaviors are chosen to be studied

1. Larger samples of participants and multiple samples of behavior can increase both internal and external validity.

Case Studies [pp117-118]

A case study provides a description of an individual.

• Usually the individual is a person, but may also be a setting.

• A psychobiography is a type of case study in which a researcher applies psychological theory to explain the life of an individual.

Case Studies…

are done when an individual possesses a particularly rare, unusual, or noteworthy condition.

provide unique data about some psychological phenomenon

and the insights gained from them may lead to other research using other methods

Ch 5: Measurement Concepts Ch 6: Observing Behavior.

Documents

reliability reliability

reliability of measures

interobserver reliability

reliability pp

test slide

content validity

testretest reliability

happy slide