Mnemonic Similarity Task - Set D

Technical Manual

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

Created: January 28, 2014

Last Modified: January 24, 2024 by K. Borchert (katjab@millisecond.com), Millisecond

Script Copyright © Millisecond Software, LLC

Background

This script implements the Mnemonic Similarity Task* with sets of 192 stimuli as described in:

Stark, S.M, Yassa, M.A., Lacy, J.W. & Stark, C.E.L. (2013). A task to assess behavioral patternseparation (BPS) in humans: Data from healthy aging and mild cognitive impairment. Neuropsychologia, 51, 2442–2449.

From the article (Stark et al, 2013): "[...] the BPS-Otask provides a sensitive measure for observing changes in memory performance across the life span and may be useful for the early detection of memory impairments that may provide an early signal of later development to mild cognitive impairment."

• the task was originally called "Behavioral Pattern Separation Task" and renamed in 2015 to "Mnemonic Similarity Task (MST)"

Background information, original programs (Win/Mac) and stimuli/instructions freely available via: https://faculty.sites.uci.edu/starklab/mnemonic-similarity-task-mst/ 2 more stimuli sets (E, F) were added to the original task.

Newest versions on: https://github.com/celstark/MST

Millisecond thanks Dr. Stark for sharing the new site with us!

NOTE: There is NO parameter to choose randomization seeds in this script as it is provided in the original. The seed used is randomized in this script.

Duration

12 minutes

Description

Participants participate in a 2-part experiment to assess recognition memory. The first part of the experiment presents 128 pictures (default) of everyday items and participants have to decide whether the item is an OUTDOOR or an INDOOR item. The second part of the experiment presents 64 of the previously seen pictures (targets), 64 of very similar items (lures), and 64 new items (foils). Participants are asked to categorize the items as old, new, or similar within 2.5s (default).

Stark et al (2013) categorized the Lures into 5 lure bins: "[...]the stimuli were divided into 5 lures bins, with the more mnemonically similar lures in lure bin1 (L1) and the least mnemonically similar lure items in lure bin5(L5) (p.2446)"

Procedure

Block Sequence: List Generations -> Part 1 -> Part 2

Lures, Targets, Foils and Stimuluspresentation (part1) List Generations:
• the stimuli are randomly divided into lures, targets, and foils lists depending on
parameters.stimSet (C or D) and selected parameters.nr_subsetImages (editable parameter)
• list.stimulusPresentation (part 1) contains the itemNumbers selected for the lures and targets
( lure lists are assembled with the constraint that each of the 5 lure bins have at least expressions.minlurefreq members)

Part 1: Stimulus Presentation (default: 128 trials => can be set via Editable Values, parameters.nr_subsetImages)
• stimuli are randomly selected from list.stimulusPresentation
• stimuli are presented for 2s (default, can be set via Editable Values)
with 0.5s interstimulus interval (default, can be set via Editable Values) => response window 2.5s
• in non-selfpaced mode (parameters.selfpaced == false): response needs to be made within 2s
• in selfpaced mode (parameters.selfpaced == true): picture is presented for 2s but participant has infinite time to make response

The task is to categorize the items as Indoor vs. Outdoor items
-> accuracy of response is NOT assessed

Part 2: Recognition Test (default: 192 trials, 64 targets, 64 lures, 64 foils => can be set via Editable Values, parameters.nr_subsetImages)
• 64 stimuli are Old stimuli from Part 1 (the same picture is presented) => Targets
• the remaining stimuli from Part 1 are Similar stimuli (a similar picture of the item is presented) => Lures
• 64 stimuli are New stimuli (not previously presented in Part 1, still from same set) => Foils
• order of stimuli is random
• stimuli are presented for 2s (default, can be set via Editable Values)
with 0.5s interstimulus interval (default, can be set via Editable Values)
• in non-selfpaced mode (parameters.selfpaced == false): response needs to be made within 2s
• in selfpaced mode (parameters.selfpaced == true): picture is presented for 2s but participant has infinite time to make response

Stimuli

There are two sets of stimuli: set C and set D. This script runs set C by default (editable parameter)
The sets are downloaded via https://darwin.bio.uci.edu/~cestark/BPSO/bpso.html (inactivated)
(now: https://faculty.sites.uci.edu/starklab/mnemonic-similarity-task-mst/)

Each set comes with a/b variants of each items:
The 'a' variants are used for part1 and for targets and foils in part2. (-> item.Stimuli/item.stimuli)
The 'b' variants are used for lures in part2. (-> item.Stimuli_lures/item.stimuli_lures)

The stimuli for each set are stored in separate subfolders.
If you only need a specific subfolder, it's not necessary to upload the second folder as well.

Instructions

Instructions are adapted from the originals downloaded via https://darwin.bio.uci.edu/~cestark/BPSO/bpso.html (inactivated)
(now: https://faculty.sites.uci.edu/starklab/mnemonic-similarity-task-mst/)
They can be easily customized under EDITABLE CODE -> Editable Instructions

Summary Data

File Name: bps_o_setc_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
showResponseKeyReminder True (1) = a text reminder is presented onscreen (while pictures are presented) that reminds participants of response keys
false (0) = no response key reminder is presented (default)
stimSet "C" or "D" (whichever set was selected)
setSize Number of trials per trial type condition in part 2 (default: 64)
nrSubsetImages Number of items per trial type condition as well as subset number (default: "64")
Choose from
"64",
"20-1", "20-2", "20-3" => sets are separated into 3 test sets for repeated measures
"32-1", "32-2" => sets are separated into 2 test sets for repeated measures
if something other than these options is chosen, the script defaults to "64"
selfpaced True = the task is self-paced => the pictures are still presented for the pre-determined duration
but participants have as much time as needed to complete the respective tasks
before the next stimulus shows up
false = task is not self-paced (default)
stimulusDuration Stimulus presentation time (default: 2000ms)
isi Interstimulus interval (default: 500ms)
countl1-
countl5 Helper variables that keep track of how many lures of each of the 5 categories have been sorted into list.lures
minLureFreq The minimum number of lures of each lure bin
countOldTargets Counts all "old" responses to targets
countSimTargets Counts all "similar" responses to targets
countNewTargets Counts all "new" responses to targets
countTargets Counts all target trials
countTargetsCorr Counts all target trials that were responded to => all target trials minus those with no responses
(same for lures and foils)
rawRateOldTargets Proportion of "old" responses to all target trials (includes no responses)
rawRateSimTargets Proportion of "similar" responses to all target trials (includes no responses)
rawRateNewTargets Proportion of "new" responses to all target trials (includes no responses)
(same for lures and foils)
corrRateOldTargets Adjusted proportion of "old" responses to target trials that were responded to (no responses excluded)
corrRateSimTargets Adjusted proportion of "similar" responses to target trials that were responded to (no responses excluded)
corrRateNewTargets Adjusted proportion of "new" responses to target trials that were responded to (no responses excluded)
(same for lures and foils)
Lure Bins Summary Statistics
l1:-L5 Counts how often lures of category 1-5 were run (raw count)
l1NR-
l5NR Counts 'no reponses' (NR) to all 5 bin lure categories
L1O -
l5O Counts how often lures of category 1-5 were categorized as "old" (raw count)
l1S-
l5S Counts how often lures of category 1-5 were categorized as "similar" (raw count)
l1N-
l5N Counts how often lures of category 1-5 were categorized as "new" (raw count)
rawPercentCorrect Overall percent correct rate (takes all responses into account; includes no responses)
corrPercentCorrect Adjusted overall percent correct rate (takes only those trials with a response into account; no responses are excluded)
mst Mnemonic Similarity Task: rates corrected for no responses
=> "hit" rate for lures (=proportion of similar responses to lure objects) - "false alarm" rate for foils (=proportion of similar responses to foil objects)
trs Traditional Recognition Score (TRS): rates corrected for no responses
=> hit rate for targets (=proportion of old responses to old objects) - false alarm rate for foils (=proportion of old responses to foil objects)

Raw Data

File Name: bps_o_setc_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 Participant ID
group Group number
session Session number
stimSet "C" or "D" (whichever set was selected)
setSize Number items per targets/foils/lures
nrSubsetImages Number of items per trial type condition as well as subset (default: "64")
Choose from
"64" => no subsets possible
"20-1", "20-2", "20-3" => sets are separated into 3 test sets for repeated measures
"32-1", "32-2" => sets are separated into 2 test sets for repeated measures
if something other than these options is chosen, the script defaults to "64"
blockCode Name of the current block
blockNum Number of the current block
trialCode Name of the current trial
trialNum Number of the current trial
stimulus Stores the currently presented picture
stimulusSelect The itemnumber of the currently presented picture
trialType The current trial type
1 = target
2 = foil
3 = lure
lureBin The lure bin of current lure (1-5); 0 for targets and foils
response The scancode of the response key
part1: 23 = I (indoor); 24 = O (outdoor)
part2; 47 = V (old); 49 = N (new); 48 = B (similar)
responseCategory The 'interpreted response'
part1; "indoor" vs. "outdoor"
part2: "old", "new", "similar"
correct The correctness of the response (1 = correct; 0 = incorrect)
only relevant for Part 2 (accuracy of responses in Part 1 is not assessed in this script)
latency The response latency in ms

Parameters

The procedure can be adjusted by setting the following parameters.

NameDescriptionDefault
showResponseKeyReminder True = a text reminder is presented onscreen (while pictures are presented) that
reminds participants of response keys
false = no response key reminder is presented (default)
nrSubsetImages Number of items per trial type condition as well as set number
Choose from
"64" => no subsets possible
"20-1", "20-2", "20-3" => sets are separated into 3 test sets for repeated measures
"32-1", "32-2" => sets are separated into 2 test sets for repeated measures
if something other than these options is chosen, the script defaults to "64"
=> sets of 32: in this script the original set(s) of 192 items was randomly divided into 2 sets of 96 items
(ensuring that each set received a minimum number of 13 lures per lure bins)
=> sets of 20: in this script the original set(s) of 192 items was randomly divided into 3 sets of 60 items
(ensuring that each set received a minimum number of 9 lures per lure bins)
All these sub sets are the same across participants (not generated dynamically for each participant)
"64"
selfpaced True = the task is self-paced => the pictures are still presented for the pre-determined duration
but participants have as much time as needed to complete the respective tasks
before the next stimulus shows up
false = task is not self-paced (default)
Duration Variables
stimulusDuration Stimulus presentation time 2000ms
isi Interstimulus interval
the combined duration of stimulusduration and ISI = response window
500ms