« Return to Top»
Method
Materials
Words and faces made up the stimuli.
There were 18 categories of words, a subset of those used by Shiffrin, Huber and Marinelli
(1995): nine categories were related by semantic similarity and nine categories were
related by orthographic-phonemic similarity. Each word category consisted of one
prototype word and 11 exemplars. The semantic categories consisted of relatively
long words of low frequency, each with a semantic link to the prototype word (e.g.
prototype: dinosaur; exemplars: fossils, extinction, reptiles, swamps, glaciers,
skeletons, brontosaurus, amphibians, mammoth, geology, artifacts).
Orthographic-phonemic categories consisted of three- or four-letter high frequency words
that shared a vowel sound and exactly one consonant cluster with the prototype word (e.g.
prototype: sip; exemplars: tip, lip, hip, sis, sin, sit, dip, rip, six, nip, pip).
There were 30 extra semantic words and 20 extra orthographic-phonemic words, with no
obvious relation to any of the categories.
There were 9 categories of female faces and 9
categories of male faces, each consisting of 12 perceptually similar exemplars, as judged
by the author. There were 20 extra female faces and 20 extra male faces, chosen not
to be similar to any of the categories (called unrelated distractors). The faces
were black and white photographic quality with similar backgrounds, selected primarily
from college yearbooks and from the Olivetti Research Database of Faces (AT&T
Laboratories, 1994). Each face was standardized so that the head orientation, level
of the eyes, and position of the chin were identical and there was very little (if any)
background. My labels for the categories are as follows. Female: old age,
middle age, young blonde straight hair, young blonde curly hair, young brown curly hair,
young brown straight hair, Asian American, African American, and Indian American.
Male: old age, middle age, young blonde hair, young with mustaches, young brown
hair, young receding hairlines, Asian American, African American, Indian American.
The extra faces came from categories above, but they all wore glasses.
Procedure
Incidental Study Session.
Participants were told that the study was
designed to investigate how people relate faces with words in advertisements; they were
not informed that a memory test would follow. On each study trial, one word appeared
directly above one face in the center of the screen for 2.5 s; participants judged the
word-face association by using the mouse to click a number on a 5-point scale. The
study session began and ended with 10 buffer trials. The buffer words were randomly
chosen from the extra semantic words and of the buffer faces, five were randomly chosen
from each of the extra female and extra male lists.
For each participant, categories were randomly
assigned lengths of two, six, or nine such that there were three categories of female
faces, male faces, semantic words, and orthographic-phonemic words assigned to each
length. The category prototypes were not presented during study. The
experimental study combinations were random pairings of faces with words, and the order of
presentation was random with respect to category membership.
Test Session.
The test session began with an explanation of
the memory test to follow. The test session consisted of two blocks, one block each
of 104 trials of single words or single faces with the order of the blocks and
presentation of items within blocks randomized for each participant. On each trial,
a single face or word appeared until the participant answered the following two
questions. First, the participants was asked to chose a number on a 6-point scale,
indicating their confidence that the exact item had appeared during the study session (1=
no confidence, 2= very little, 3= little, 4= moderate, 5= high, 6= very high). The
second question asked the participants to use the keypad to give their best estimate of
the number of items on the study list that were similar to the test item in appearance,
sound, spelling, or meaning, including the test item itself, if it had been studied.
Each test block began with four buffer items
chosen from the remaining extra items. The buffer items were two female and two male
faces for the face block, and four extra semantic words for the word block.
For each word category, the prototype, two
studied words, and two non studied words were tested. The test block also included
six of the remaining extra semantic words and four of the extra orthographic-phonemic
words. For each face category, two studied faces and three non studied faces were
tested. The test block also included five of the extra male faces, and five of the
extra female faces. After the buffer items, the test items within block were
presented in random order.
« Return to
Top»
Last revised:June 03, 2003 11:46:58 AM
Copyright© 1995 The University
of Mississippi. All rights reserved.
Questions about this page?PsychExps