Stan Cronk, PhD; Pardha Parimi, BS
Center for Biomedical Engineering and Rehabilitation Science
Louisiana Tech University
Ruston, LA 71270

ABSTRACT

Engineering students are expected to be able to estimate readings from scales and gauges to one significant digit beyond the smallest scale graduation. We conducted a set of experiments to compare the tactile length estimation skills of blind students with the visual and tactile estimation skills of sighted students. We found that persons who are blind performed comparably to sighted individuals who performed the tests using only their sense of touch, but were slower and less accurate than persons using sight.

KEYWORDS:

blind, tactile, estimation skills

BACKGROUND

When reading instruments such as scales or gauges, one must normally must estimate quantities that fall between the smallest graduations on the instrument. As a rule, sighted individuals should be able to estimate one additional significant digit of a measurement beyond the smallest graduation on a ruler or gauge (1).

As engineering programs in universities continue to work to make our programs accessible to all individuals, we want to ensure that our laboratories and measurement instruments provide equal access. We have been unable to find any previous research that examined whether persons who are blind have the same length estimation skills as persons who are sighted, though Smith et al (2) determined that persons who are blind have a superior ability to persons who are sighted in manually recalling lengths of objects which the individuals had previously examined by touch.

RESEARCH QUESTION

The objective of this study was to compare the tactile length estimation skills of persons who are blind with the tactile and visual estimation skills of persons who have the use of sight.

METHODOLOGY

The scale of the graph is from 0 to 1 units. The graph shows a horizontal line along the x-axis that is a length 80% of the graph scale.(Click image for larger view)

We prepared a set of simple full-page graphs in Microsoft EXCEL, with the line length varying horizontally along the x-axis. We printed each graph first on a laser printer. We then ran each printed graph through a Repro-tronics Tactile Image Enhancer along with a sheet of Flexi-Paper. The Repro-tronics unit used heat to print a raised image on the Flexi-Paper corresponding to each part of the printed graph containing a line or character. Figure 1 shows a sample graph. On half the sheets, the scale of the x-axis ranged from 0 to 1, with these points marked in Braille at the left and right ends of the axis. On the other half of the sheets, , the scale of the x-axis ranged from 0 to 2, with these points marked in Braille at the left and right ends of the axis, and point “1” \marked in Braille halfway in between points “0” and “2”. We prepared 18 graphs with line distances ranging from 0.1 to 0.9 units in increments of 0.1 units. The graphs were prepared to be printed as a full page, so the length of “0.1 units” was 10.4 mm for graphs in which the x-axis ranged from 0 to 2 units, and 20.9 mm for graphs in which the x-axis ranged from 0 to 1 unit.

Twenty-four persons with ages ranging from 15 to 25 participated in the experiment. Eight participants were blind, while 16 persons had sight. Six of the eight blind participants were in high school, while two had graduated from college. All 16 of the sighted participants were college freshmen majoring in engineering. Two of the blind participants were female, while two of the sighted participants were female.

We explained to each individual what was required during the estimation task, and informed each person that the lengths they were measuring would range from 0.1 to 0.9 units in increments of 0.1 units. We also informed each participant that some lengths could be repeated during the tests. We asked each individual performing the tests tactilely to explore a sample page to orient themselves to the key characteristics of each graph, principally the starting and ending points of the lines as well as the scale of the graph (0 to 1 or 0 to 2 units).

**Table 1. Sample data collected for one participant.**
Sheet	Test #	Actual Length	Guess
A	4	0.5	0.5
B	8	0.8	0.8
C	2	0.2	0.2
D	3	0.2	0.8
E	17	0.9	0.9
F	14	0.4	0.5
G	15	0.1	0.1
H	10	0.8	0.8
I	7	0.1	0.1
J	12	0.3	0.3
K	16	0.6	0.7
L	1	0.7	0.6
M	6	0.3	0.3
N	11	0.4	0.5
O	18	0.6	0.8
P	5	0.7	0.8
Q	13	0.5	0.7
R	9	0.9	0.9

The participants who were blind only performed the tactile tests. Half of the participants with sight performed the set of estimation tasks first using their sight, while half performed the set of tasks first with their sense of touch. Each time a person used touch to estimate distance, the investigator asked that participant first to show the endpoints of the line and to state if the maximum line distance was one unit or two units. The purpose of this was to ensure that the participant was basing their estimation on an accurate interpretation of the graph.

RESULTS

Table 1 shows a sample set of data collected for one subject. The column Sheet is the code we used to name the graph; sheets A through I contained scales of 0 to 2, while sheets J through R contained graphs with scales of 0 to 1. The column “Test #” shows the order in which the tests were actually conducted. Columns “Actual Length” and “Guess” are self-explanatory.

Table 2 summarizes some of the performance data we collected.

**Table 2. Summary data**
Summary Data	Blind	Sighted using Vision	Sighted using Touch
Mean Time to complete tests (sec)	711	212	1235
Mean Percentage of Tasks Correct	50%	89.60%	50%
Range of Errors	0 – 14	0 – 7	4 – 11
Mean Magnitude of Error	0.15	0.06	0.14

DISCUSSION

We found that persons who are blind were much slower to complete the length estimation tasks than persons who were using their vision (711 seconds vs. 212 seconds), and were not as accurate in estimating lengths as persons using vision (50.0% of estimations correct vs. 89.6% correct for persons using vision). However, the time required to complete the tests was much less than that of sighted persons performing the tests tactilely. The accuracy of persons who were blind was essentially equal to those of sighted individuals performing the estimation tasks tactilely.

The mean magnitude of error for sighted individuals was 0.06 units, somewhat smaller than the 1 significant digit of error expected. Though the mean error for blind participants was 0.15 units, a magnitude 2.5 times greater than that of sighted individuals, the error was only 50% greater than one significant digit.

The results of these tests seem to indicate that a person who is blind would be able to estimate the last significant digit of a measurement reasonably well, if an analog scale can be presented in a tactile format. However, these tests were conducted with somewhat large differences between graduations on the scale (10.4 mm or 20.9 mm), requiring movement of the forearm to feel the endpoints of the line. An additional set of tests using smaller scale graduations, similar in size to analog scales used in engineering labs, should be conducted to support our conclusion.

REFERENCES

Eide, Arvid; Jenison, Roland; Mashaw, Lane; and Northup, Larry. (2002). Engineering Fundamentals and Problem Solving, Fourth Edition. McGraw-Hill Higher Education Publishing. 184-185.
Smith, Melissa; Franz, Elizabeth; Joy, Susan; Whitehead, Kirsty. (2005). Superior Performance of Blind Compared with Sighted Individuals on Bimanual Estimations of Object Size. Psychological Science 16(1): 11-14.

Stan Cronk, PhD
Center for Biomedical Engineering and Rehabilitation Science
Louisiana Tech University
711 S. Vienna
Ruston, LA 71270
(318) 257-4562; cronk@latech.edu

A Comparison of Length Estimation Skills between Blind and Sighted Individuals

Stan Cronk, PhD; Pardha Parimi, BS
Center for Biomedical Engineering and Rehabilitation Science
Louisiana Tech University
Ruston, LA 71270

ABSTRACT

KEYWORDS:

BACKGROUND

RESEARCH QUESTION

METHODOLOGY

RESULTS

DISCUSSION

REFERENCES

Highlights

Gargoyles

Sheet	Test #	Actual Length	Guess
A	4	0.5	0.5
B	8	0.8	0.8
C	2	0.2	0.2
D	3	0.2	0.8
E	17	0.9	0.9
F	14	0.4	0.5
G	15	0.1	0.1
H	10	0.8	0.8
I	7	0.1	0.1
J	12	0.3	0.3
K	16	0.6	0.7
L	1	0.7	0.6
M	6	0.3	0.3
N	11	0.4	0.5
O	18	0.6	0.8
P	5	0.7	0.8
Q	13	0.5	0.7
R	9	0.9	0.9

Sheet	Test #	Actual Length	Guess
A	4	0.5	0.5
B	8	0.8	0.8
C	2	0.2	0.2
D	3	0.2	0.8
E	17	0.9	0.9
F	14	0.4	0.5
G	15	0.1	0.1
H	10	0.8	0.8
I	7	0.1	0.1
J	12	0.3	0.3
K	16	0.6	0.7
L	1	0.7	0.6
M	6	0.3	0.3
N	11	0.4	0.5
O	18	0.6	0.8
P	5	0.7	0.8
Q	13	0.5	0.7
R	9	0.9	0.9

A Comparison of Length Estimation Skills between Blind and Sighted Individuals

Stan Cronk, PhD; Pardha Parimi, BS Center for Biomedical Engineering and Rehabilitation Science Louisiana Tech University Ruston, LA 71270

ABSTRACT

KEYWORDS:

BACKGROUND

RESEARCH QUESTION

METHODOLOGY

RESULTS

DISCUSSION

REFERENCES

Highlights

Gargoyles

Stan Cronk, PhD; Pardha Parimi, BS
Center for Biomedical Engineering and Rehabilitation Science
Louisiana Tech University
Ruston, LA 71270

Sheet	Test #	Actual Length	Guess
A	4	0.5	0.5
B	8	0.8	0.8
C	2	0.2	0.2
D	3	0.2	0.8
E	17	0.9	0.9
F	14	0.4	0.5
G	15	0.1	0.1
H	10	0.8	0.8
I	7	0.1	0.1
J	12	0.3	0.3
K	16	0.6	0.7
L	1	0.7	0.6
M	6	0.3	0.3
N	11	0.4	0.5
O	18	0.6	0.8
P	5	0.7	0.8
Q	13	0.5	0.7
R	9	0.9	0.9