Research Methods

• 1. to solve practical problems - applied research
• 2. To satisfy curiosity and gain scientific understanding - Basic Research

• -Motivation: SPECIFIC need
• -Purpose:knowledge applied to a very specific situation
• -Application: narrow

• -Motivation:curiosity-desire for explanation
• -Purpose:basic principles
• -application: broad

• 1. theoretical constructs (concepts or conceptual variables likely not directly observable)
• 2. State (make educated guess)
• a. relationship between theoretical
• constructs OR
• b. that one theoretical construct
• causes another
• Ex: violent TV content - aggression
• Dreaming in color - creative
• Bumper stickers - Road Rage

convert abstract concepts (theoretical constructs) into specific (testable) variables

-has a range of values (levels)

mass X acceleration

mass X gravitational field strength

1.18 X US dollar

Gender

continuous numerical scale

Discrete

• -Independent Variable
• -Dependent Variable
• -Control Variable

E changes (manipulates) (possible cause)

first in time: independent of the results of the study.

one value of an I.V.

An independent variable has at least 2 levels of a treatment. If not- NOT a variable

measured by E (possible effect)

Second in time: depends on I.V.

could be independent variable but held constant (not the same as control group or condition)

• -Predictor variable
• -Outcome or Criterion variable

• No manipulation
• Possible cause; first in time

Possible EFFECT; second in time

D.V. ; C.V.

Between subjects comparison

within subjects comparison

Exact description of procedure used to generate independent variable: how to produce each level of the independent variable

Exact description of how to "obtain" dependent variable: how to measure or recognize a particular thing or characteristic

Statement describing what to do (produce, create, construct, generate) different amounts or levels of the variable

Example: fear vs no fear / frown vs no frown (categories)

statement of what to do to measure (or how to measure) a particular thing or characteristic

ex: fear; smile vs frown; neuroticism

• -Clear, precise, objective...so repeatable by others
• -Captures at least some part of the concept you are trying to measure (valid)
• -Can be done consistently (reliably); clarifies the meaning of the concept

Practical

I.V. - Cartoons with violence and Cartoons without violence (Describe criteria for violence)

D.V. - Number of times child hits BoBo doll (criteria for hit?)

• 1. public verification
• 2. Blueprint for experiment
• fear causes thigmotaxis
• agreed upon methodology, systemati
• observation
• 3. Evaluation of research
• captures ideas that researcher
• claims to test. quality of research

• gas, pressure, temperature
• crowds and helping
• noise and performance
• weight loss and activity

it explains a variety of facts

poor predictive power

Develop a number of different hypotheses/theories

design and conduct "critical tests" of each

Gather (1) falsifying and (2) confirming (supporting) evidence

Lots of supporting evidence and little or no falsifying evidence and closest to the truth

(at the moment)

Collect specific observations to infer general principles

logically valid

"if-then"

Consequent: there were more fights

• 1. If P, then Q
• 2. Q
• 3. Therefore P

• If smith is a mother, then she is female
• Smith is a female
• Therefore smith is a mother

always possible new data will be collected

Logically valid

• if p then q
• not q
• therefore not p

the act of characterizing observations

1. Develop o.d. - clear procedures for measuring or classifying observations and producing variables: publicly verifiable, reliable

a.)Decide in advance what you are going to measure(esp. important in naturalistic ob.)

b.)decide how to assign numbers (or categories) to observations

tend to notice and look for evidence that confirms expectations

Collect data

Summariza data

Data from individuals not representative

a.) measures of central tendancy

• Mode
• Median
• Arithmetic mean
• Geometric mean

All measures of variability show how much scores from the standard, baseline or reference

the average of the squared distances each score is from the mean

Take the square root: standard deviation

• 1. find mean
• 2. subtract mean from each score
• 3. square resulting number
• 4. add together
• 5. divide by N

proportional to the average distance a set or scores is from the mean

• Sensitivity
• Reliability
• validity of a test/measurement

increases with the number of possible values that can be consistently used.

• differentiate
• discriminate
• distinguish

stability orconsistency-sometimes over time; repeatable, reproducible

precise in sense that random, UNSYSTEMATIC errors of measurement are minimal/nonexistent

Obtained Score = true score + unsystematic measurement error

(look at many scores from many individuals)

variability of scores - reduces precision

balance out.

use arithmetic mean

equal 0

"true" variance

• Systematic variance
• _________________
• systematic+error variance

• Total Variance
• (will get a number between 0 and 1)

Consistency of a measure from one time to the next

• Unsystematic errors - cancel out
• Systematic errors - do NOT cancel out
• Unsystematic errors - influence variablity
• Systematic errors - influence mean - (poor variability)

• Time:influenced by time between tests
• Change
• Carry over

Consistency of the results of a test constructed in the same way and from the same content area.

consistency of results across items within a test designed to measure same construct.

• 1. Average inter-item correlation
• 2. Split-half reliability

Average of all possible split half reliability

degree to which different raters or observers give consistent estimates of the same phenomenon

• it is supposed to
• it is designed to
• purports to measure

extent to which it uses items or content representative of the area (concept) you are trying to measure

Construction of test--theory and research

• 1.Items truly from content area
• 2.Items are representative same of
• content area

Degree to which a test correlates with some direct and independent measure of what you are trying to measure.

• IQ
• boss's ratings of salesmanship
• weights? crimes?

a single criterion/operational definition

Future

Same time

Extent to which a test or measure can be shown to measure a particular theoretical construct-conceptual variable (unobservable abstract trait or feature)

• 1. test produces "numbers" distinct from that produced by a measure of another construct
• 2. Based on accumulated evidence

characteristic or trait(construct) to be measured-theoretical relationships etc.

Sensation seeking: a trait describing the tendancy to seek novel, varied, complex, and intense sensations and experiences and the willingness to take risks for the sake of the experience

Correlate test with a variety of measures that should be positively, negatively, or not correlated with characteristic (construct)

Examine the pattern of results using a diverse body of evidence.

• Convergent validity
• Discriminent Validity
• Criterion Validity

strong cor. between test and conceptually similar measures

low cor. between test and measures of different theoretical constructs

strong cor. between test and direct and independent measure (ex. behavior)

that test is correlated with itself-reproducible

• internal validity
• external validity

degree to which a study measures the effect of hypothesized cause or independent variable

-Did different levels of the treatment (IV) cause the change in the outcome(DV)?

Degree to which the finings can be generalized to other subjects, and to other situations (settings, levels of variables, ways of measurement etc)

yea-saying (or the opposite: response deviation--nay saying)

• ecological validity
• population validity

the degree to which sample scores can be generalized to the target population

tend to overestimate the extent to which other people share or behaviors, attitudes, and beliefs

the degree to which the independent and dependent variables accurately reflect or measure what they are intended to.

the extent to which one can generalize from the research setting and participant population to other settings and populations

refers to whether one can make causal statements about the relationship between variables

refers to the consistency of behavioral measures

involves dividing the test items into 2 arbitrary groups and correlating the scores obtained in the 2 halves of the test

divides the population into smaller units

occurs when a statistical test can detect effects

reletively objective technique for summarizing across many studies investigating a single topic

example: lunar-lunacy hypothesis. full moon makes for more aggressin

scaling of concepts such as brightness

applies when concepts, such as depression, are measured but usually do not have clearly specified inputs

For a particular sensory modality, the size of the difference threshold relative to the standard stimulus is a constant.

A popular way of assessing psychological traits that do notseem to lie on a known physical scale.

Provides a score for a psychometric property of a person that derives from how that person responds to several statements about a topic that clearly are favorable or unfavorable

is the standard deviation of a distribution of sample means