Single Factor Designs

Single Factor Designs

Between-Subjects Between Subjects, Two Levels
• Independent Groups
– random a do assg assignment, e t, many a y pa participants tcpa ts
– e.g., typical experiments in social Psychology
– independent groups t-test
• Matched Groups
– smaller number of participants
– some factor strongly correlates with DV
– e.g., sleep deprivation and suggestibility study
– with one-to-one matching, use dependent groups t-test
• Nonequivalent groups
– subject variable
– match the groups on other characteristics
• bbut t not t necessarily il a one-to-one t matching thi
– e.g., neuropsychology

Within-subjects, Two Levels
• once per condition, use complete counterbalancing
• twice per condition, use reverse counterbalancing
– e.g., Stroop (1935)
• more than twice, use block randomization
– sometimes can use alternation (Lee and Aronson, 1974)

• Single Factor, Multi-level
– Discover non-linear effects
• Ebbinghaus percent savings
• u-shaped u shaped dose curve
– multiple control conditions
• Mozart effect
– multiple control groups
• Bransford and Johnson (1972)

If the balloons popped, the sounds
wouldn’t be able to carry since
everything would be too far away
ffrom the th correct t floor. fl A closed l d
window would also prevent the
sound from carrying, since most
buildings tend to be well insulated insulated.
Since the whole operation depends
on a steady flow of electricity, a
break in the middle of the wire
would also cause problems. Of
course, the fellow could shout, but
the human voice is not loud enough
to carry that far. far An additional
problem is that the string could
break on the instrument. Then
there could be no accompaniment
to the message. It is clear that the
best situation would involve less
distance. Then there would be
fewer potential problems. With
face-to-face contact, the least
number of things could go wrong.

• Presenting Data
– In the text
– Tables
• Exact values
– Graphs
• IV on X, DV on Y

• Types of graphs
– Line graphs (continuous IV -- interpolation)
• More than 2 levels
– Bar graphs (discrete IV -- categorical)
• or continuous with just two levels (immediate versus 24 hours)
– Scatter plots (correlation)
– Error bars (whenever possible!)

Multiple Pairwise comparisons
• If there are more than 2 conditions, you may
want to make more than one simple comparison
(more than one t test)
• what’s the level for 3 conditions with 3
comparisons? What about J comparisons?
• Wh What t about b t dependencies d d i bbetween t comparisons? i ?
• Bonferroni inequality gives upper limit to observing one or
more type I errors (≤J)

Control group designs
– Placebo control groups
– Waiting list control groups
• can be combined with placebo p
– Merikle and Skanes (1992)
– Yoked control groups
• Eye y Movement Reprocessing g and Desensitization
– Dunn et al. (1996)

which test?
Within Subjects Between Subjects
matching
2 levels 3 or more levels 2 levels 3 or more levels
dependent
Repeated independent
independent
groups
measures groups
groups
tt-test test
ANOVA tt-test test
ANOVA

IVs (between, within, manipulated), DVs, design, scale (nominal,
ordinal, interval, ratio), presentation (table, graph), analysis
• College students in a cognitive mapping study are asked to use a
direction finder to point accurately to three unseen locations that
differ in distance form the laboratory laboratory. One is a nearby campus
location, one is a nearby city, and the third is a distant city.
• Three groups of preschoolers (50 per group, assigned randomly) are
iin a study t d of f ttask k perseverance in i which hi h ddelay l of f reward d is i varied. i d
The children in all three groups are given a difficult puzzle and told
to work on it as along as they would like. One groups is told that
they will receive a chocolate bar at the end of the session if they
complete l t the th puzzle. l Th The second d group will ill get t a chocolate h l t bbar after ft
20 minutes from the end of the session, and the third will get the
chocolate after 40 minutes.
• To examine whether crowding affects problem-solving performance,
participants are placed in either a large or a small room while
attempting to solve a set of word puzzles. Before assigning
participants to the two conditions conditions, the researcher takes a measure of
their verbal intelligence to ensure that the average verbal IQ of the
groups is equivalent.