PSC 041 - UC Davis Unit 2

types of variable relationships

positive linear, negative linear, no relationship

how can you tell if a variable is continuous/categorical

continuous predictor variable will be an AVERAGE

categorical predictor variable will be a PERCENT

T/F error bars do not overlap means there is a RELATIONSHIP?

TRUE REAL RELATIONSHIP

what is p-value

probability that the null hypothesis is true, evidence is stat significant if p<0.05

when would you use a scatterplot

both predictor and outcome CONTINUOUS

curvilinear relationship means

some areas of predictor and outcome pos and neg

horizontal line of best fit

no relationship

correlation coefficient (r)

A single number, ranging from -1.0 to 1.0, that indicates the strength and direction of an association between two variables.

does r value ever go above -1 or 1

NO

correlational research vs correlation coefficient

correlational research - ASSOCIATIVE, predictor NOT manipulated, variables continuous OR categorical, NO CAUSATION

correlation coefficient - usually both variables continuous, can be correlational OR experimental, CAUSAL OR ASSOCIATIVE

what is a construct

A construct is an unobservable thing (like memory or intelligence) that is useful for explaining or describing behavior.

OFTEN HAVE A HARD TIME MEASURING

what is validity

the degree to which a test measures what it says it measures.

what is reliability

the consistency with which the same event is repeatedly measured (repeat and replicate)

does good reliability = good validity?

NO

how do you calculate reliability

correlation coefficient (r)

tests to check reliability

test and retest (same participants)

parallel forms (new participants)

split half (single test administered, compare odd and even questions)

observation measures

action

self report measures

ideas

monitoring measures

physiological state

self report vs observation

self report - report feelings after the fact

observation - recording feelings while event occurring

qualitative data

subjective response (feelings)

quantitative data

objective response (one right answer) COMPARED TO RUBRIC

open ended question

an interview question that encourages an answer phrased in the respondent's own words

closed ended question (3 types)

multiple choice (forced choice, exhaustive)

rating scale questions (how much agrees)

ranking scale (arrange options in order)

response set

a tendency to respond to questions in a particular way that is unrelated to the content of the questions

nay saying

answering "no" or "strongly disagree" to every item in a questionnaire or interview

yea saying

pattern may be evident in the form of all "yes" or "strongly agree" answers.

reverse wording

changes the direction of the scale by asking the question in a positive (or negative) voice

faking bad

Giving answers on a survey (or other self-report measure) that make one look worse than one really is.

no causation without

manipulation

four types of validity

construct, face, method, procedural

face validity

Measures whether a test looks like it tests what it is supposed to test.

procedural validity

did anything about the procedure add noise or error to the measurement?

is validity formulaic

NO

method validity

did researchers appropriately measure variables in experiment

construct (face validity) is an example of

THINGS WE CANT SEE

procedural/method validity is example of

THINGS WE CAN SEE

method match is only for

measured (not manipulated) variable

values of r (0.50 to 1.00)

strong relationship

values of r (0.30 to 0.50)

moderate relationship

values of r (0.10 to 0.30)

weak relationship

values of r (0.00 to 0.10)

no relationship

r for

relationship!

p for

probability!

direction of relationship (- or +)

(-) negative linear relationship aka inverse

(+) positive linear relationship

does r value ever go above 1?

NO

for calculating reliability what value do you use?

correlation coefficient (r)

r goes with R!

reliability r = +1 to +0.9

reliability excellent

reliability r = +0.9 to +0.8

reliability good

reliability r = +0.8 to +0.7

reliability acceptable

construct validity pro tips

did they actually measure what they said?

make evaluation

use sound reasoning

leave out details, leave out outcome

if a likert scale has 6+ points it is

continuous

do error bars need to be the same height?

no

threats to construct validity

is construct really measured?

is operational definition valid?

is there external factors (observer expectancy, participant expectancy, hawthorne effect, faking bad/social desirability)

line of best fit

A line that is very close to most of the data points in a scatter plot

How to write an operational definition

Identify the characteristic of interest.

Identify the characteristic to be measured or the defect type of concern.

Select the measuring instrument.

The measuring instrument is usually either a physical piece of measuring equipment such as a micrometer, weighing scale, or clock; or alternatively, a visual check. Whenever a visual check is used, it is necessary to state whether normal eyesight is to be used or a visual aid such as a magnifying glass. In the example, normal eyesight is sufficient. In the example, a clear visual indication is given of acceptable and unacceptable, so the observer needs to be in a position where the decision can be made.

Describe the test method.

The test method is the actual procedure used for taking the measurement. When measuring time, the start and finish points of the test need to be specified. When taking any measurement, the degree of accuracy also needs to be stated. For instance, it is important to know whether time will be measured in hours, minutes, or seconds.

State the decision criteria.

The decision criteria represents the conclusion of the test. Does the problem exist? Is the item correct? Whenever a visual check is used, a clear definition of acceptable/unacceptable