1 of 16

How we know

What we know

2 of 16

Research Fundamentals

Experimental Research

Descriptive Research

variables manipulated & recorded

variables recorded

variables in an around the expt. are controlled

variables in and around the study not controlled

participants are (randomly) assigned to groups

participants not assigned to any groups

variables include IV, DV, EV, confounds

variables include predictor & predicted

stats used to detect differences between means

stats used to detect relationships among scores

strength = cause & effect

strength = understanding relationships

weakness = uncertain generalizability

weakness = confounds, no cause & effect

3 of 16

Research Fundamentals

VARIABLES: What experiments are made of

Independent: the variable manipulated by the experimenter to see if the � manipulation affects some other variable.

“manipulated” means creating different groups or conditions

to be compared

Does teaching method affect _____? Create groups to compare online vs F-2-F

Does type of test affect ______? Create groups to compare open vs closed book

Does a teacher’s sex affect _____? Create groups to compare male vs female

Does vape juice flavor affect _____? Create groups to compare none, fruit, complex

4 of 16

Research Fundamentals

VARIABLES: What experiments are made of

Dependent: the variable affected by the IV…. that is, values of the DV will depend on which group or condition of the IV the participant was in.

Does teaching method affect test scores? See if test scores differ between online vs F-2-F�

Does type of test affect test anxiety? See if anxiety scores differ between open vs closed book�

Does a teacher’s sex affect student evals? See if eval scores differ between male vs female�

Does vape juice flavor affect daily puffs? See if # daily puffs differ b/w none, fruit, complex

5 of 16

Research Fundamentals

VARIABLES: What experiments are made of

Extraneous: any variable other than the IV or DV in an experiment

teaching method

(test scores)

type of test

(anxiety levels)

teacher’s sex

(student eval)

vape juice flavor

(# puffs per day)

sex of teacher

sex of student

sex of student

sex of vaper

sex of student

pre-existing anxiety levels

teaching method

vape device

GPA

GPA

grade in class

vape cloud

prof’s experience

testing location

teacher’s age

years vaping

FR, SO, JR, SR

time of day

FR, SO, JR, SR

time of day

6 of 16

Research Fundamentals

Confound: an extraneous variable that is systematically different between the groups, so that the experimenter no longer knows if it was the IV or the extraneous variable that produced the effect on the DV

teaching method

sex of teacher

sex of student

GPA

prof’s experience

FR, SO, JR, SR

teaching method

online

Face-2-Face

male

female

x x

x x

x x

mean mean

DV = test scores

7 of 16

Research Fundamentals

Confound: an extraneous variable that is systematically different between the groups, so that the experimenter no longer knows if it was the IV or the extraneous variable that produced the effect on the DV

type of test

open book

closed book

low GPA

high GPA

x x

x x

x x

mean mean

type of test

sex of student

pre-existing anxiety levels

GPA

testing location

time of day

DV= test anxiety

8 of 16

Research Fundamentals

Confound: an extraneous variable that is systematically different between the groups, so that the experimenter no longer knows if it was the IV or the extraneous variable that produced the effect on the DV

teacher sex

male

female

C or lower

B or higher

x x

x x

x x

mean mean

teacher’s sex

sex of student

teaching method

grade in class

teacher’s age

FR, SO, JR, SR

DV = student evals.

9 of 16

Research Fundamentals

Confound: an extraneous variable that is systematically different between the groups, so that the experimenter no longer knows if it was the IV or the extraneous variable that produced the effect on the DV

vape juice flavor

x x x

x x x

x x x

mean mean mean

vape juice flavor

sex of vaper

vape device

vape cloud

years vaping

time of day

none

fruit

complex

small

medium

large

DV = # puffs / day

10 of 16

Research Fundamentals

DESCRIPTIVE RESEARCH

Correlational Research

Two or more variables are observed, measured, and recorded to

see if they are related to each other.

The experimenter does not manipulate (create groups or conditions) for any of the variables.

He/she simply observes and records what’s already there.

Is the number of scars on the body related to age?

Is the number of years spent smoking related to years lived?

Are IQ scores related to annual income?

11 of 16

Research Fundamentals

DESCRIPTIVE RESEARCH

Correlational Research: STRENGTH

Typically, the data are analyzed with a statistic called the “correlation coefficient” “r

-1 ---------------- 0 --------------- +1

Perfect No Perfect

Relationship Relationship Relationship

  • The larger the r value (+ or -) the stronger the relationship

r = .77 r = .47 r = - .81

12 of 16

Research Fundamentals

DESCRIPTIVE RESEARCH

Correlational Research: DIRECTION

+r means as x increases, y increases OR

as x decreases, y decreases

-r means as x increases, y decreases OR

as x decreases, y increases

Is the number of scars on the body related to age?

Is the number of years spent smoking related to years lived?

Are IQ scores related to annual income?

r is positive

r is negative

r is positive

13 of 16

Correlation is NOT Causation

Even if the relationship the variables is perfect, this does not mean one “caused” the other!

X might cause Y

Y might cause X

a third variable might be the true cause

14 of 16

MEDIATING AND MODERATING VARIABLES

Mediator: a variable that explains why or how the IV had an effect on the DV

or why two or more variables are related (as in correlation)

Texting while driving increases driving errors

Why or how?

Texting is a distraction and distracted drivers make more errors

Distraction is the mediator

15 of 16

Moderator a variable that qualifies the effect of an IV on the DV in an � experiment by answering questions like “for whom”, “when”, or

“under what circumstances”. Moderators can also qualify the � relationship between variables as in correlation.

it is the “yes but” variable

Texting while driving increases driving errors

BUT

This is effect is greater for city driving than highway driving

16 of 16

Statistical Significance

Statistically Significant, p < .05 What does it mean?

The effect, difference, or relationship found is most likely a TRUE

effect, difference, or relationship

The probability that the effect, difference, or relationship is just

a chance finding is less than 5%

Just because a result is significant does NOT

mean it is important!