Experimental essay

Bill 1
False Memory

In our study, we will test whether short-term and long-term tests show different levels of
false memory for different kinds of lists. We will be running two different tests which included a
total of 370 participants in the University. In experiment one
Introduction
Our research questions:
1. Can we replicate McBride et al. showing more STM false memories for phonological
than semantic lists?
2. Do STM and LTM show different list effects on false memories (i.e., an interaction
between test delay and type of list?
Independent variable 1 (IV1): List Type (phonological and semantic)
Independent variable 2 (IV2): Test Delay (short and long)
Dependent variable (DV): false memory (measured by the proportion of “yes” responses to
critical lures)
Please run two male subjects and two female participants through the following procedure. Run
the subjects one at a time. The subject’s participation is voluntary. All subjects must be 18 years
old or older. Please follow all ethical guidelines in the treatment of your subjects. The
experiment will take less than 20 min for each subject. If a subject wishes to discontinue the
experiment at any time, you must allow them to do so. If this happens, you will need to replace
that subject in the experiment. Make sure each subject signs the consent form. You may not run
any subject until they sign the consent form.
Bill 3
Study of false memories for as long as there has been memory research
• Roediger and McDermott (1995) popularized a method for experimentally inducing simple
false memories
• Themed lists are studied, but theme (aka critical lure) not studied
• Participants show false memories for the lures
• Roediger and McDermott (1995) showed both false recognition and false recall for the critical
lures
• Recall rates as high as accurate recall of list items in the middle of the lists
Method
Participants
● 370 total participants ran in the experiment.
● 339 sets of data were analyzed
● 30 sets of data were deleted due to experimenter error
● 1 set of data was deleted due to failure to follow instructions
● All were 18 years or older
● Half were men, half were women
Design
Materials
Bill 4
Be sure you have a timer ready for the distractor task (state capitals). The task sheets the subjects
will complete are included below. Prepare these pages before you run the experiment. You can
email them to the participants if run online or print out if run in person.
After the subject signs the consent form, read the instructions printed below out loud and follow
the procedure in order. Ask the subject if they have any questions and if they do, answer them
before they begin. Then show the PPT with the study lists and test items to the subject. Be sure
to have the distractor task ready before you start the long delay test.
Procedure
“In this experiment, I will ask you to view a timed PPT with 4-item lists of words shown.
After each list, you will see a + sign. Then you will either see a word to decide if it was in the list
you just saw or an instruction to circle “no”. If a word is shown, please try to remember if it was
in the list of 4 words that was just shown. Circle “yes” if you think it was and “no” if you think it
was not. At the end of the lists, the instructions for the next task will appear and I will read those
to you.
Bill 5
Results
Overall means (marginal means):
*Short Delay M = .44
*Long Delay M = .77
Semantic List M = .59
Phonological List M = .62
DV was number of lures participants indicated were studied
● Test was a within-subjects (or repeated-measures) ANOVA
● Alpha level for significance was .05
● Main Effect of Delay was significant:
F(1,338) = 66.06, p < .001 (p is less than .05)
● Main Effect of List Type was not significant:
F(1,338) = 1.09, p = .30 (p is greater than .05)
● Interaction between Delay and List Type was not significant:
F(1,338) = 2.85, p = .09 (p is greater than .05)
● At short delay, more FMs for phono lists than semantic lists:
Bill 6
F(1,338) = 5.88, p = .02 (p is less than .05)
● At long delay, no difference in FMs across lists:
F(1,338) = 0.16, p = .69 (p is greater than .05)
● Overall, there were more false memories at the long delay than the short delay. (state this
with means and stat values included)
● Although the interaction between delay and list type was not significant, follow-up tests
showed more false memories for phonological lists than semantic lists at the short delay.
This result replicates the results reported by McBride et al. (2019).
Bill 7
Discussion
Hypothesis 1: supported – we found that there were more false memories for
phonological than semantic lists at the short delay – this replicates the McBride et al. (2019)
findings
● Hypothesis 2: only partly supported – the short delay showed a difference across lists, but
the long delay did not
Bill 8
Bill 9