Deese–Roediger–McDermott - DRM False Memory Procedure - Visual

Technical Manual

Script Author: Katja Borchert, Ph.D. (katjab@millisecond.com), Millisecond

Created: January 27, 2016

Last Modified: January 13, 2023 by K. Borchert (katjab@millisecond.com), Millisecond

Script Copyright © Millisecond Software, LLC

Background

This script implements the Deese–Roediger–McDermott (DRM) procedure to study false memories based on:

Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology, 58, 17–22.

Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in list. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803–814. (Experiment 2)

Note: this script presents the study words visually.

Duration

60 minutes

Description

Participants study 16 lists of 15 related words. The words are presented visually with a Stimulus Onset Asynchrony (SOA) of 1500ms. One item that is highly related to all the words of one list ('lure') is not included in the study lists. In an immediate recall test after 8 randomly selected lists, participants have to list as many of the study words as they recall. Number of correctly recalled items as well number of times (and position on recall list) of the 'lure' item is assessed by computer. Participants work on simple addition and multiplication problems after the remaining 8 lists (=control task). Once all 16 lists have been studied, participants are given a recognition test (old/new) of 96 randomly presented items and remember-know judgments if the item has been identified as "old". The recognition test contains all 'lure' items of the study lists (16) as well as 3 study items of each of the 16 study lists (48). The remaining items are new items.

Procedure

- 24 lists of 15 study items + 1 'lure' item (=16 items per list) get randomly assigned to be used for study-recall lists (8),
study-math lists (8) and recognition-only lists (8)

(I) Study-Phase:
- the order of study-recall and study-math tasks (8 each) is randomly determined
- the presentation order of the study words during the study phase is sequential
- words are presented visually with a SOA 1500ms

Recall Task:
Algorithm to estimate number of recalled words:
In general this algorithm uses the number of a specific word separator (here: comma)
to estimate the number of words in between the separators: estimated word count = number of separators + 1
The algorithm uses several steps to clean up the text input to ensure as much as
possible that only single commas are used as word separators. It also removes
trailing commas from the input if necessary.

Steps:
(1) textbox response is stored in values.Recall (Example: values.Recall = "bed; chair; table") (note: there is a space AND a ; after bed and chair)
(2) values.Recall replaces 7 types of possible word separator symbols (e.g. ";", " ") with a comma (Example: values.Recall = "bed,,chair,,table")
(3) values.Recall adds a comma to the first item (Example: values.Recall = ",bed,,chair,,table")
(3) values.Recall replaces all consecutive commas with a single comma (Example: values.Recall = ",bed,chair,table")
(4) if values.Recall does not end in a comma at this point, this comma is added (Example: values.Recall = ",bed,chair,table,")
(5) the length of the variable values.Recall is stored in variable values.recallLength (Example: values.recallLength = 17)
(6) a second variable (values.reducedRecall) deletes all commas in values.Recall (Example: values.reducedRecall = "bedchairtable" => length: 13)
(7) the word count estimate = length of values.Recall* - length of values.reducedRecall - 1 (extra comma) (Example: estimate = 17 - 13 - 1 = 3)
•with length = number of characters

the recall scoring algorithms implemented in this script depend on participants
(a) using the implemented separators
(b) make no spelling mistakes
(c) don't use phonetically similar words (e.g. pain instead of pane)
Manual checks of the raw data might be necessary to adjust the scores (see values.intrustionOther)


(II) Recognition Phase:
- 96 items (48 old/48 new): order is randomly determined
48 => 3 from all 16 study lists (items selected: position1, 8, 10) (old words)
24 => the 'lures' of all 24 lists (new words)
24 => 3 from the 8 recognition only study lists (items selected: position1, 8, 10) (new words)
- old/new judgments
- if old: remember-know judgments

Stimuli

- words published by Roediger & McDermott (1995)
- study words presented black on white in the center of the screen (size can be adjusted under section Editable Parameters)

Instructions

instructions are provided by Millisecond and followed the guidelines outlined in
Roediger & McDermott (1995). The Instructions can be edited under section Editable Instructions.

Summary Data

File Name: falsememories_visual_summary*.iqdat

Data Fields

NameDescription
inquisit.version Inquisit version number
computer.platform Device platform: win | mac |ios | android
startDate Date the session was run
startTime Time the session was run
subjectId Participant ID
groupId Group number
sessionId Session number
elapsedTime Session duration in ms
completed 0 = Test was not completed
1 = Test was completed
meanWordRecall Mean of the estimated number of words recalled
the word count estimation is done by computer using the algorithm described below.
There may be instances when the algorithm does not return the correct count.
Furthermore, the algorithm does NOT check if the input consists of actual words.
sdWordRecall Standard deviation of the estimated number of words recalled
propWord1Recall-
propWord15Recall Stores the proportion of correctly recalled words in study positions 1 - 12 (Max: 6/6 for each item)/propLureRecall: stores proportion of recalling the lure (=a specific highly related item that was NOT a study item) based on the number of recall trials (Max: 6/6)
spelling errors of correctly recalled items are counted as 'errors' by the search algorithm
(search algorithm only searches for correctly spelled words)
propLureRecall Stores recall proportion of the lure (=a specific highly related item that was NOT a study item) based on the number of recall trials (Max: 8/8)
meanLurePosition Calculates the mean lure Position in the list of recalled items across all lists
sdLurePosition Standard deviation of the lure positions in the list of recalled items across all lists
meanIntrusionsOther Calculates the mean number of other word intrusions on recall
! a value > 0 could be a flag to manually check in the raw data file (e.g. values.intrusionOther) whether misspellings of study list words
occurred as these misspelled word would be counted as 'other intrusions'
recognition
propHits Hit rate (across all studied words)
propHitsRecallWords Hit rate for study words used in recall blocks
propHitsMathWords Hit rate for study words used in math blocks
propFA False alarm rate (across all non-studied words)
propFARecallLures False alarm rate for lures tied to study words in recall blocks
propFAMathLures False alarm rate for lures tied to study words in math blocks
propFANewWords False alarm rate for study words from lists that were not studied
propFANewLures False alarm rate for lures of non-studied lists
propRememberRecallWords Proportion remember judgments for hits for study words used in recall blocks
propRememberRecallLures Proportion remember judgments for false alarms for recall study word lures
propRememberMathWords Proportion remember judgments for hits for study words used in math blocks
propRememberMathLures Proportion remember judgments for false alarms for math study word lures
propRememberNewWords Proportion remember judgments for false alarms for study words from lists that were not studied
propRememberNewLures Proportion remember judgments for false alarms for lures of non-studied lists
math
nrMathProblems Number of math problems attempted
propCorrectMath Proportion correct math problems

Raw Data

File Name: falsememories_visual_raw*.iqdat

Data Fields

NameDescription
build Inquisit version number
computer.platform Device platform: win | mac |ios | android
date Date the session was run
time Time the session was run
subject, group, With the current subject/groupnumber
session Session number
blockCode Name of the current block
blockNum Number of the current block
trialCode Name of the current trial
trialNum Number of the current trial
listCount Stores the number of lists run
recallListCount Counts number of recall lists run
mathListCount Counts number of math lists run
nextList Stores the list index (1-15 of the currently presented study list)
lure Stores the lure of the currently presented study list
wordCategory Recognition Phase: "new", "new-lure", "recall" (for study words used with recall task), "recall-lure", "math" (for study words used with math task), "math-lure"
stimulusItem The presented stimuli in order of trial presentation
response The participant's response
correct The correctness of the response (1 = correct; 0 = incorrect) - if applicable
latency The response latency (in ms)
recall Stores the recall text input (with single commas to separate the different words)
countCorrectRecall Tracks the number of correctly recalled words per list
recallWordEstimate Stores the computer-derived estimate of the number of words recalled
see below of algoritm implemented. Algorithm does not check whether the recalled words are actual words.
positionLureRecall If lure is present in recall: stores the position of the lure in the recall list
nrOtherRecalledWords Stores the number of other recalled words that are neither study words nor lures
intrusionOther Stores the character strings contained in the recall string that are neither the study words nor the lure
misspelled study words will show up here

Parameters

The procedure can be adjusted by setting the following parameters.

NameDescriptionDefault
wordPresentationSOA The stimulus onset asynchrony (in ms) of the presented study words 1500ms
recallDelay The delay (in ms) between last presented study list word and the recall trial 0ms
maxRecallDuration The time (in ms) allotted to the recall of the words 2.5mins or 150000ms
minRecognitionDelay The minimum delay (in ms) between end of last trial and the recognition task 2 min or 120000ms
studyWordSize Proportional sizing of study words 5% of canvas height