Instructional Illustrations & Graphical Devices:
Designer’s Intentions & Readers’ Interpretations
Elizabeth Boling
Kennon Smith
Theodore Frick
Indiana University
Malinda Eccarius
University of Nebraska Lincoln
c 2004 DO NOT REPRODUCE WITHOUT PERMISSION
Illustrations are widely used in textbooks
º Illustrations are widely used in instructional materials
(Evans, Watson, and Willows,1987; Pettersson, 2002),
constituting some 30 – 60% of page space in junior high
texts and up to 80% of page space in elementary texts.
º Pictorial images have demonstrated potentially
beneficial effects in a broad range of settings, and for a
variety of types of learning (Anglin, Towers & Levie,
1996).
Graphical devices are used in many of instructional
illustrations to extend the meaning of the pictorial image. In
some cases the student must interpret the device in the way
the designer of the image intended in order to complete a
learning task successfully.
Accurate interpretation of instructional
illustrations is not a given.
º Despite what appears to largely be a cross-cultural
ability to recognize objects depicted in pictures (Kennedy,
1994; Sless, 1981), the visual content of an illustration is
frequently a vehicle used to communicate a more
complex meaning or intention.
º Beyond their ability to present a visual representation of
a given object, visual illustrations do not constitute a
universal language.
Research questions
º To what extent do various populations interpret the
meaning of simple illustrations including graphical devices
consistently with the meaning intended by the designer of
the illustrations?
º When interpretations are not consistent with the intention
of the designer, what are the ways in which respondents
interpret the meanings of these illustrations?
Pilot Study
º 96 college students at a large Midwestern university
completed a paper-based survey containing sixteen
illustration items.
º For each item, each participant wrote a short (usually
one sentence) response, indicating what they believed the
illustration to mean.
º Based on this pilot study, the images were redrawn for
consistency, and a 2nd version of the survey was produced
without graphical devices in order to collect comparison
data.
Survey items
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
path of prior motion
communication / source of
communication
expression (happy, excited)
physical change (sick, dead)
Examples from a Vietnamese text for
learning English
property of motion (fast)
Zwier, L. (2003). English for everyday conversation and activities.
Examples from a Vietnamese text for
learning English
property of communication
(electronic)
Zwier, L. (2003). English for everyday conversation and activities.
Examples from a Vietnamese text for
learning English
communication
Expression (dismay,
unhappiness)
Zwier, L. (2003). English for everyday conversation and activities.
Examples from a Vietnamese text for
communication
learning English
communication
motion
Zwier, L. (2003). English for everyday conversation and activities.
Examples from a Vietnamese text for
learning English
motion
Zwier, L. (2003). English for everyday conversation and activities.
Examples from a Vietnamese text for
learning English
motion
Zwier, L. (2003). English for everyday conversation and activities.
motion &
path of
motion
Survey populations
Surveys with
devices
Surveys without
devices
US 3rd graders
38
25
US 6th graders
46
39
US 10th graders
26
28
US college students
34
39
Malay college
students
50
54
US teachers of the
deaf & hard of
hearing
47
45
Totals
241
230
Coding
Matches the designer’s intention
1
Does not match the designer’s
intention
0
There’s a lizard = 0
Dead lizard = 1
There’s a lizard = 0
Dead lizard = 1
Results
º People trained on the coding scheme agree very
consistently in their coding.
º The mean of the Kappa values (simple percentage
agreement corrected for chance agreement) is .89, with a
standard deviation of .15
º Surveys returned in languages other than English were
coded by native speakers of those languages.
Frequency analysis
3rd grade
N=38
6th grade
N=46
10thgrade
N=26
US college
N=34
Teachers
N=47
Malay
N=50
bunny
0.87
0.83
0.85
0.74
0.98
0.56
flower
0.13
0.28
0.19
0.38
0.47
0.34
car
0.58
0.74
0.85
0.74
0.72
0.72
gift
0.18
0.44
0.38
0.24
0.43
0.52
running guy
0.13
0.26
0.38
0.21
0.45
0.10
hat
0.40
0.56
0.65
0.82
0.85
0.66
baseball
0.50
0.59
0.62
0.56
0.72
0.22
talking
0.97
0.98
0.96
0.94
1.00
0.84
sleep/cook
0.03
0.02
0.00
0.00
0.00
0.00
no running
0.79
0.94
1.00
1.00
1.00
0.94
signing
0.58
0.67
0.81
0.71
0.89
0.38
transform
0.71
0.89
0.81
0.65
0.47
0.18
TV
0.40
0.61
0.69
0.76
0.83
0.52
looking
0.79
0.85
0.85
0.91
0.94
0.64
dog
0.90
0.83
0.88
0.82
0.94
0.58
lizard
0.63
0.80
0.81
0.68
0.53
0.42
Frequencies for all populations by image, with graphical devices
Frequency analysis
3rd grade
N=38
6th grade
N=46
10thgrade
N=26
US college
N=34
Teachers
N=47
Malay
N=50
bunny
0.80
0.85
0.86
0.77
0.96
0.44
flower
0.04
0.00
0.00
0.00
0.02
0.00
car
0.16
0.36
0.18
0.13
0.24
0.26
gift
0.00
0.00
0.00
0.00
0.00
0.04
running guy
0.00
0.00
0.00
0.00
0.00
0.00
hat
0.16
0.00
0.00
0.08
0.07
0.15
baseball
0.36
0.46
0.79
0.62
0.69
0.13
talking
0.40
0.49
0.75
0.82
0.87
0.54
sleep/cook
0.00
0.00
0.00
0.00
0.00
0.00
no running
0.00
0.00
0.00
0.00
0.02
0.02
signing
0.52
0.72
0.61
0.80
0.96
0.57
transform
0.00
0.00
0.00
0.00
0.00
0.00
TV
0.02
0.18
0.18
0.15
0.33
0.18
looking
0.04
0.13
0.18
0.23
0.29
0.13
dog
0.00
0.00
0.07
0.00
0.02
0.04
lizard
0.08
0.13
0.14
0.05
0.09
0.18
Frequencies for all populations by image, without graphical devices
Results of frequency analysis
º Graphical devices do make a difference in the
way that respondents interpret the illustrations:
º Respondents who did not see the devices answered
consistently with designers’ intentions infrequently
except for the images with redundant cues to
meaning (bunny, talking, fast car, looking, sign
language)
º 2 out of 3 who see the devices match the designer’s
intended meaning; 1 out of 5 who do not see the
devices match the designer’s intended meaning
Results of frequency analysis
º Respondents do not interpret meanings as
consistently as designers might want or expect
them to:
º Only two images are interpreted correctly at 80% or
above across populations (no running, verbal
communication - two people talking)
º 10th graders interpreted 9 of 16 pictures consistently
with the designers’ intentions at 80% or above;
teachers of the deaf interpreted 8 out of 16 pictures
this way
º In all other populations fewer than half the pictures
are interpreted correctly at 80% or above
Comparative analysis
º Because the data violate the homogeneity of variance
assumption, standard ANOVA procedures could not be
used to compare how different groups responded to any
particular image.
º Therefore, a z-test procedure was used to compare the
frequencies with which groups interpreted given images
consistently with the designer’s intention.
z-test comparisons
º For each image, we made pairwise comparisons with an
overall alpha level of .05
º For each image, four pairwise comparisons were
calculated. Therefore, the overall alpha level of .05 was
divided by 4, resulting in an alpha of .0125 for each
individual comparison.
º This process is conservative, and minimizes the probability
of committing type 1 errors (finding differences between
groups simply by chance).
Sample comparative analysis - item 13
designer’s intended meaning
electronic speech
group
frequency
3rd graders
0.39
10th graders
0.69
Teachers
0.83
US college
0.76
Malay college
0.44
-2.431 (nsd)
-3.390 (sd)
-0.705 (nsd)
3.172 (sd)
t-test values (equal variance not assumed)
3rd grade & 10th
grade students
3rd grade & US
college students
US college & Malay
college students
US college &
Teachers of deaf
0.244
1.404
0.736
-3.055
-0.630
-2.478
0.391
-0.765
-2.460
-1.399
0.153
0.117
-1.723
-0.524
-2.772
-1.833
-2.258
-0.829
1.285
-2.371
-2.081
-4.114
1.725
-0.325
-0.906
-0.493
3.235
-1.514
0.261
0.668
2.239
-1.436
1.000
1.000
******
******
-3.141
-3.141
1.769
******
-2.022
-1.118
3.093
-2.054
-0.896
0.568
4.686
1.612
-2.431
-3.390
3.172
-0.705
-0.576
-1.469
3.216
-0.399
0.124
0.854
2.515
-1.492
-1.575
-0.395
2.381
1.317
Comparative analysis
º The greatest number of significant differences were found
between the U.S. college students and the Malay college
students.
º Few differences were found between U.S. college students
and U.S. teachers of the deaf and hard of hearing, and
few differences were found between U.S. 3rd grade and
U.S. 10th grade students.
º U.S. college students are more similar in their
performance to U.S. 3rd graders than to Malay college
students.
º The data to date suggest that developmental factors
make less difference in interpreting images than do
culture and language differences. (In this case, the
designer of the images was from the U.S.)
Qualitative analysis
U.S. college students – all responses, with and without devices
Qualitative analysis
U.S. 3rd graders – all responses, with and without devices
Qualitative analysis
º Using the constant comparison method to sort all
responses made by US college students, we have
identified three major groupings of answers.
1 … form of the response
all answers from all respondents –
those who saw the graphical
devices and those who did not
2 … interpretation of the device
answers from respondents who
saw the devices but did not
match the intended meaning
3 … interpretation of the image
Answers from respondents who
did not see the devices and did
not match the intended meaning
Findings
Form of the response (all responses)
code
definition of the code
example
NR
no response
---
N
naming
lizard
D
description
The lizard layed there
eN
embellished narrative
The lizard got hit by a car, and
died.
Ne
narrative element
I am a lizard
C
commentary
I have a pet lizard.
Findings
Interpreting the device (images with devices; answers do no match
the designer’s intention)
code
definition of the code
example
D
ascribes a meaning to the device
without reference to the relationship
between the device and the rest of
the image
The wind is blowing.
R
interprets the device with a different
relationship to the image than
intended by the designer
The flower is being
pushed by wind
/D
interprets the image in a way that
would be possible if the device were
not present
The flower is wilting
/M
matches the interpretation of the
device but not the intended meaning
of the illustration
The flower is leaning the
wrong way.
Findings
Interpreting the image (images without devices; answers do no
match the designer’s intention)
code
definition of the code
example
M
matches the intention of
the designer without the
device present
A rabbit is jumping.
N
names or describes the
noun
a rabbit
C
interpreting a
The rabbit is falling.
characteristic of the image
with a different result
than intended
/C
unclear cue or
interpretation
There he goes.
Findings
º Respondents seeing the images without devices are 5
times more likely than those who see the devices to give
a descriptive response. The devices make a qualitative
difference in the form of the responses.
º Respondents whose interpretations do not match those of
the designer give interpretations that are related to the
devices. They appear to notice and use the devices in
making their interpretations rather than ignoring the
devices or not noticing them. Designers cannot assume
that people will ignore graphical devices that they do not
understand.
Limitations of the study
º Images are simple and viewed outside the learning
context.
º Investigators do not have access to the respondents’
reasoning process for giving the answers they do.
º Images are not categorized by type, so within-subject
analysis is not possible at this time.
Note (02/28/05):
An expanded version and double-checked
version of this study has been accepted
for publication in the Journal of Visual
Literacy. Please do not reproduce or
circulate data or text from this
presentation.
Descargar

Slide 1