Braille Reading Rates as a Function of Reading Tasks

M. Knowlton, R. Wetzel

Abstract: This study of the cognitive processes of braille reading compared the reading of 23 adult braille readers in four different reading conditions: oral reading, silent reading, studying, and scanning. The findings provide support for the idea that braille reading is process driven and that reading rates vary, depending on the purpose of the reading task.

Much has been written about the psychophysical phenomena of braille reading. Some researchers have investigated issues related to the nature of the braille stimulus, such as size and spacing, character recognition, and reading rate, and others have examined the best educational practice in teaching braille: characteristics of braille readers, teaching techniques, and methods to facilitate the reading process. However, reading comprehension has received much less attention. The aim of the study of the reading behavior of expert braille readers presented in this article was to move research in the field from a primarily psychophysical and perceptual level to a more cognitive level. Thus, the focus was on comprehension and reading rates as the purpose of reading tasks is manipulated. Some of the research questions were as follows:

  1. Are reading rates of 80-120 words per minute (wpm) an accurate measure of the reading rates of adult readers who use braille materials daily?
  2. Do reading rates change when the demands of various tasks change?
  3. What effects do scanning, studying, or silent reading have on an individual's braille reading rate?

Review of the literature

A review of the pertinent research on these issues provides a background for the study that follows. This section presents a summary of research specifically on braille reading, identifies current issues in the process of reading, and describes a model of the reading process proposed by Carver (1981), called "rauding."

BRAILLE READING

Significant research has been conducted on braille reading in the past 90 years. Commencing with the studies of the Uniform Type Committee (1908, 1910, 1912, 1913, 1915) extensive research has been conducted on the optimal structure of the braille cell, including size, number of dots, dot placement, and use of contractions. As a result of the committee's work, 63 braille characters are used today. Additional early research similarly focused on the psychophysics of the stimuli (Burklen, 1932; Meyers, Ethington, & Ashcroft, 1958) or on the presentation rates for braille (Flanigan & Joslin, 1969; Nolan & Kederis, 1969).

Studies by Foulke (1964) and Nolan and Kederis (1969) placed the average oral reading rate at 90 wpm using students in Grades 5-12 who read materials appropriate to their grade levels. Two studies reported exceptions to this rate. The first study, by Lowenfeld, Abel, and Hatlen (1969), found reading rates of 116 wpm for students in residential programs and of 149 wpm for students in regular local schools. The second study, by Grunwald (1966), observed reading rates of 200-300 wpm. Although demographic information on the subjects in Grunwald's study were not reported, Nolan and Kederis attributed what they called the "spuriously high reading rates" in both studies to the sampling procedures that were used and reiterated their position that braille is fundamentally a slow way to read.

A different interpretation of these results was implied by Foulke (1991, p. 227), who suggested that a different perceptual focus is required for rapid braille readers: "The stream of dots that flows beneath the reading fingers varies in spatial extension as a function of time. These variations form memorable patterns with which syllables, words and even phrases can be associated."

Other researchers have gone beyond the psychophysics of braille to examine issues related to learning and reading braille (Newman, Hall, Foster, & Gupta, 1984; Newman et al., 1982) and different strategies used by good and poor braille readers (Millar, 1977). Because of a concern about the relative slowness of braille reading, numerous studies have been focused on reading rates. These studies have investigated the implications of manipulating the braille code (Lorimer & Tobin, 1979; Martin & Sheffield, 1976), reading strategies (Crandell & Wallace, 1974; McBride, 1974; Olson, 1976; Olson, Harlow, & Williams, 1975, 1977), and optimal methods of teaching braille (Bims, 1976; Caton & Bradley, 1978-79). In addition, tactile discrimination and hand movements have received a significant amount of attention (Mangold, 1978; Olson et al., 1975, 1977; Wormsley, 1981). There is a consensus in the field that to lead productive lifestyles, braille readers must develop adequate skills (CEARSVH, 1991; Rex, Sowell, Koenig, Sveen, & Cheadle, 1991 ).

In reviewing the extensive research on reading braille, one sees a significant emphasis on the mechanics and perception of braille stimuli. However, there has been little, if any, research on the process of reading-of taking in symbolic information, assigning meaning to the information, and comprehending the information-which is the primary purpose of reading.

THE READING PROCESS

According to Smith (1988), the components of the reading process for sighted people include 1) perception, 2) memory, 3) identification of letters, 4) identification of words, and 5) comprehension. Although the ultimate goal of gaining access to information is the same for both sighted readers and readers with visual impairments, one must question if the reading process is the same. Is reading a different task for persons with visual impairment than it is for sighted persons? Does Smith's list of components accurately describe the reading process for readers with visual impairments, or should the components be different?

Perceptual processes

It is in looking at the perceptual processes involved in reading that one may see the greatest degree of difference between readers with visual impairments and readers with normal vision. The visual process for print readers with nominal vision is well understood (Smith, 1988). For braille readers, there is a significant amount of data on the recognizability of braille characters and braille words. Specific differences in the perceptual processes involved in the reading of braille as compared to print are beyond the scope of this paper.

Memory processes

In examining the role that memory plays in the reading process, one must look at both the sensory memory store and short-term memory (STM). The primary factor in processing information from either area is attention. Attention is required to capture information from the sensory memory store, and attention is required to maintain information in STM. This information has been well documented in the field of psychology; a reasonable discussion can be found in any introductory psychology text (see, for example, Wieten, 1989).

When evaluating the memory process with regard to braille readers, one must consider several factors. First, the individual may direct so much attention toward the tactile process that he or she uses no attention to maintain the information in memory. Second, the perceptual process may be so slow that integration of information into comprehensive units may be impossible, as in the case of students who are trying to master the intricacies of the braille code or those with some form of neuropathy that interferes with their recognition of characters. Third, the perceptual window for braille is small, especially for those who read with only one hand. As a result, the reading speed may be so greatly reduced that information is dropped from STM because of the lack of attention or other resource limitations. The processes involved in processing tactile information are assumed to be comparable to those involved in processing visual information once the information has reached STM, but there is little information to substantiate this assumption.

Two studies have investigated the relationship between the spatial-graphemic structure and the phonological structure of braille. Both Millar (1975) and Pick, Thomas, and Pick (1966) studied the effects of encoding from a spatial as well as a phonological perspective. Although the methods and goals of these studies were different, both studies concluded that phonological information enhances the performance of various tasks. Attention, in the form of rehearsal or elaboration, is required both to maintain the information in STM and to move it to long-term memory.

Letter recognition

Any discussion of the processes involved in recognizing letters includes the process involved in detecting features and matching templates. An in-depth discussion of information-processing theories is beyond the scope of this article. Rather, the question that is relevant here is: Is the stimulus-recognition process the same when one is looking at tactile stimuli? Although the early research on braille included many studies on recognizing characters, these studies focused primarily on accuracy and rate, not on the internal information-processing model.

Currently, there is no experimentally derived list of the critical features of braille reading. The closest approximation may be Nolan and Kederis's (1969) rank ordering of the recognizability of braille characters, but that list was not based on empirical research. The lack of an identified list of critical features may be due, in part, to the fact that braille characters and words are not entirely unique: The same punctographic symbols are used in literary braille, mathematics (the Nemeth Code), and music braille. A braille character may represent a single letter, a combination of letters (such as st, er and ch), or an entire word, depending on the graphemic context in which it is used. In addition, one braille character may take on different meanings when used in musical notation or in mathematics. Therefore, any discussion of the critical features of braille must be context specific.

Word recognition

With regard to the processes involved in recognizing words, three possible approaches are usually discussed: letter-by-letter recognition, part-word recognition, and whole-word recognition. Each approach is based, in part, on feature-detection theory-the concept that a critical feature or piece of a stimulus will allow one to eliminate all but the correct stimulus. Once again, the question arises: What process is involved in tactile readers' recognition of words? Although the literature on braille reading is replete with research on word recognition, the research has been based on the premise that the recognition of words in braille is a character-by-character process. There is no exact character-by-character correspondence between braille and print. Braille is a highly contracted system of symbols with specific rules for part-word and whole-word contractions. The rules for using part-word contractions are specific and may lend credence to the part-word theory of word recognition, even for braille.

Nolan and Kederis (1969) is largely responsible for the character-by-character view of braille reading. In an extensive series of studies, these researchers concluded that slowness in recognizing words in braille is a function of the need to identify words character by character. In relation to the reading process, however, this conclusion is misleading. The shorthand nature of braille permits each braille sign to represent more than a single character in print reading. The role of phonics in the identification of words, especially in teaching reading, has been hotly debated since the 1960s. In her review of research on reading and the psychological research related to reading, Adams (1990) concluded that although phonics is beneficial in teaching reading in languages in which there is a direct correspondence between symbols and sounds, the arbitrariness of English makes the usefulness of phonics to beginning readers less certain. Adams recommended that phonics should be taught as part of a comprehensive reading program that mixes both the wholeword and phonetic approach to reading.

Adams's position is similar to Smith's (1988). Smith stated that phonics can be used as a mediational tool in reading; however, since the same meaning can be obtained directly from the printed word, subvocalization is frequently a barrier to efficient reading because it limits the rate at which text can be read. In addition, it interposes an additional layer of processing that is difficult for beginning readers. Nothing in the literature to date has indicated that this situation is any less true for braille readers, but the reading rate that Foulke (1964) reported for braille reading may be slow enough to allow for subvocalization with no further decrement in reading rate.

Reading comprehension

Finally, the goal of reading is comprehension-obtaining meaning from words found on the printed page. Yet words on the printed page are just symbols; the reader must supply the meaning. Major problems in reading occur when a person has no information or experience in memory to associate with the visual information (Smith, 1988). That is one reason why information for experts can be written in a few words, whereas information for novices requires many more words: Works written for novices must attempt to provide the information that these persons do not have in memory. The problem is the same for a reader who uses print and one who uses braille. After decoding the symbols or sounds available to him or her, a reader must attach meaning to them because the symbols or sounds are meaningless in themselves. How much more difficult is this task for a visually impaired child? Does the lack of certain types of information or experience play a significant role in the comprehension of written materials?

RAUDING

Carver (1981, 1990) proposed an alternative interpretation of reading that may result in a simpler view of the reading process. He defined reading as the process of looking at visually presented words, letters, or other language symbols to gain information or knowledge. Although this definition, in itself, may not be readily applied to reading braille, Carver took a task-orientation approach to the reading process that may be useful in looking at the reading task for visually impaired readers. He defined skimming-scanning, studying, and memorizing, which are different tasks, as follows:

- Skimming-scanning: a reading process that involves looking at each individual word simply to recognize it.

- Studying: a task that involves the basic reading process, possibly supplemented by productive activities, such as notetaking, reciting, or outlining.

- Memorizing: a series of steps performed on the words of a passage to increase the probability that the facts, words, ideas, or thoughts can be freely recalled on subsequent occasions; it includes the components of lexical access, semantic encoding, sentential integration, remembering ideas, and recalling facts.

Carver viewed these tasks as different from what is traditionally thought of as reading, yet they all certainly have reading embedded in them. To describe the average reading that is done to understand information, such as reading a newspaper or a novel, Carter coined the term rauding-a combination of reading and auding. He defined rauding as "the reading rate at which an individual can comprehend 75% of the information presented to [him or her]" (p. 194).

The rauding rate is defined not only in wpm, but in sentences per minute and thoughts per minute. It is self-reported with an individual indicating the percentage of material he or she comprehended, since the emphasis is on the comprehension of the ideas presented not just on the processing of a certain number of words.

This view of reading is consistent with Smith's (1988) view that the goal of reading is comprehension, which is influenced by both the information the reader already has and his or her intent. For example, a reader will read differently when he or she is looking up a favorite television show in TV Guide (scanning), reading a new Dick Francis novel (rauding), and studying for an organic chemistry final (studying).

Rauding may be a more appropriate view of reading by persons with visual impairments, since the emphasis in this approach is not on process, but on comprehension-the ultimate goal of reading. The study reported next tested this hypothesis.

Method

The goal of the study was to investigate the effects of various reading tasks on the reading performance of expert adult braille readers.

SUBJECTS

The 23 subjects who participated in the study were either totally blind or had limited useful vision, and all but one used braille as their primary reading mode from the time they entered school at ages 5-7; the remaining subject did not begin braille instruction until age 9 (for demographic information on the subjects, see Table 1). These adult braille readers were recruited through the local chapter of the National Federation of the Blind and from the community on the basis of their known skill as braille readers.

Table 1 Demographic information on the subjects.
Subject Number Age Sex College Education (Years) Age at which taught to read braille Preferred reading hand Braille reading (hours per week)
1 40 M 3 6 Left 4-5
2 45 F 4 7 Right 35-42
3 41 M 4 6 Right 21-28
4 65 F 4 6 * *
5 43 M 4 6 Left 3-4
6 40 M 6 6 Right 14-21
7 68 F 0 5 Right 3-4
8 23 M 6 9 Right 28-35
9 63 M 0 6 Right 28-35
10 44 F 4 6 * *
11 38 F 6 6 Right 5-6
12 37 M 3 6 Right 35-42
13 49 M 5 6 Left 35-42
14 45 F 5 6 Right 28-35
15 49 M 8 6 Left 35-42
16 45 F 4 6 Right 21-28
17 29 M 4 6 Right 14-21
18 27 M 2 6 Right 42-49
19 46 F 5 5 Right 42-56
20 48 F 0 6 Right 18-20
21 31 F 7 7 Right 35-42
22 50 F 6 6 Right 14-21
23 30 F 5 6 Right 18-20

*Data for preferred hand and hours spent reading are not available for these subjects because of an examiner's error. However, the data on reading rates were retained in the analysis because these two variables were not included in this study.

TASK

The subjects were asked to read under four different reading conditions:

1. oral reading: a decoding task in which the subjects read aloud as fast as possible with no concern for comprehension.

2. silent reading: a modified rauding task in which the subjects were required to read and understand the materials.

3. studying: a reading task in which the subjects were allowed unlimited time to comprehend and retain the material.

4. scanning: a recognition task during which the subjects identified specific information.

In the experimental sessions, the subjects read a short story aloud and then read multiple samples of braille under the three remaining reading conditions: silent reading, studying, and scanning in one of three random orders. At the end of the sessions the subjects completed the vocabulary portion of the Weschler Adult Intelligence Scale (WAIS), which was administered to verify that they had adequate verbal comprehension levels for the reading tasks.

The study was designed as a 4 x 3 factorial design. There were two within-group factors'. 1) the reading task, with four levels (oral reading, silent reading, studying, and scanning'), and 2) presentation order, with three levels. Except for the oral reading task, which was always presented first, the other three tasks were presented in a Latin square design in which the subjects were assigned to one of three presentation orders for the remaining tasks: 1) silent reading, studying, and scanning; 2) studying, scanning, and silent reading; or 3) scanning, silent reading, and studying.

MATERIALS

Reading materials for this study were selected on the basis of their probable unfamiliarity to the subjects. The following is a list of reading materials by task:

1 . Oral reading: "The Empress of the Sea" (Rosen, 1991)

2. Silent reading: excerpts from In the Shade of the City: A History of the Loring Park Neighborhood (Trimble, 199 1)

3. Studying: excerpts from Physics Made Simple (Freeman & Dunden, 1990)

4. Scanning: Recipes from Vegetarian Ti*es (July 1991 issue).

The two investigators generated questions for both the studying task and the silent reading task to provide an evaluative measure of comprehension. For each of the studying tasks, each investigator independently developed five questions. Initial agreement on the content of the two sets of questions was 87 percent, computed by dividing the total number of questions agreed on by the total number of questions generated. Questions on which there was no initial agreement were discussed, and a set of questions were selected on which there was agreement. A true-false format was chosen for all but one of the questions.

Questions for the silent reading tasks were developed on the basis of a textual analysis of each excerpt. Each investigator parsed each sentence into the number of thoughts or facts it contained. For example, the sentence, "The more people who go through the Greenway, the better, the more exciting, the safer, and the more interesting it will be" (Trimble, 1991, p. 128) was broken into the following thoughts:

  1. The more people who go through the Greenway, the better it will be.
  2. The more people who go through the Greenway, the more exciting it will be.
  3. The more people who go through the Greenway, the safer it will be.
  4. The more people who go through the Greenway, the more interesting it will be.

The proportion of agreement by the two investigators on the number and content of the thoughts contained was 92 percent (obtained by dividing the number of agreed-on thoughts by the total number of thoughts generated). Questions were then derived turning the thought around and asking it as a question; for example, the statement "Westminster Presbyterian Church was started in 1857" became the question "What year was Westminster Presbyterian Church started?"

PROCEDURE

The subjects were assigned to one of the three presentation-order conditions and then were asked to provide some basic demographic information: age, number of years of formal education, and age at which they began to read. For the first reading task, oral reading, each subject was instructed to read the passage aloud once, as rapidly as possible, and to read for speed, not for feeling or effect. The second, third, and fourth tasks were dependent on which condition the subjects were assigned to.

Silent reading task. The silent reading task was a modified reading task. The subjects were instructed to read the article through only once, as rapidly as possible, and were told that immediately afterward, they would be asked to give a narrative account of what they had read and then would be asked questions to aid their recall. Of the three trials, only the second and third were scored; the first trial was considered a practice trial to familiarize the subjects with the task.

Studying task. The subjects were told that they could take as much time as they needed for this task and to indicate when they were finished. They were informed in advance that they would be asked several questions on the text following their study of it. Of the three tasks, only the second and third were scored; the first was considered a practice task.

Scanning task. The subjects were informed that they would be asked a question and that their task would be to find the correct answer. Of the four tasks, only the second, third, and fourth were scored; the first was considered a practice task.

Vocabulary portion of the WAIS. This subtest was administered to measure the verbal skills of the subjects apart from the reading task.

DATA COLLECTION

All the experimental sessions were audiotaped and the tapes were timed after the experimental sessions to reduce pressure on the readers and to ensure the accuracy of the recorded data. The following data were collected for each reading task:

1. Oral reading: number of wpms; the first minute of reading time was excluded for each subject as a warm-up time.

2.Silent reading: number of minutes reading, number of wpms, and number of correct responses.

3.Studying: number of minutes studying, number of wpms, and number of correct responses.

4.Scanning: number of wpms.

In the scanning task, the computation of the number of wpms was based on the number of words up to the correct answer, on the assumption that the subjects were reading straight through the text. This, however, may not be the case because individuals may skip around in the scanning process. No meaningful measure of comprehension for the total passage was obtained, since the subjects were searching for specific information.

Results

An initial analysis performed to determine if there were any order effects for the sample found no significant effects. Table 2 presents the descriptive statistics for the four reading tasks. As you can see, the mean reading rates were as follows: oral reading, 135.9 wpm; silent reading, 104.7 wpm; studying, 105.8 wpm; and scanning, 202.9 wpm.

Reading comprehension was measured under two task conditions: studying and silent reading. For the studying task the measure of comprehension was the percentage of correctly answered questions; the mean percentage of correct answers was 82.8 (SD=15.22). For the silent reading task, the measure of comprehension was the percentage of correctly recalled details, which was computed by summing the number of correctly recalled details from the narrative report and the number of correct responses to the prompting questiori and dividing the total by the total number of questions asked. The mean percentage correct (25.2, SD=1.43)-a low comprehension rate-is a function of modifications in the rauding process that were made by the investigators to ensure comprehension of the text.

Table 2: Descriptive statistics for the four reading tasks.
Reading Task Mean SD n Minimum Maximum Range
Oral Reading 135.9 29.8 22 64.94 185.29 120.36
Silent Reading* 104.7 31.0 21 50.39 162.70 112.31
Studying* 105.8 41.2 21 54.60 240.98 112.31
Scanning** 202.9 116.2 21 70.51 561.07 490.56

*Averaged over two trials. **Averaged over three trials.

The analysis of variance to determine the effect of the four reading tasks on reading rates found that the task had a highly significant effect on the reading rates (*-13.96, p<.000*; see Table 3) A post hoc analysis indicated that there were significant differences in the reading rates for oral reading and studying but not for silent reading and scanning (see Table 4). That is, although there were no significant differences between the reading rates for silent reading and studying, the reading rates for these conditions differed significantly from the reading rates for oral reading and scanning.

Table 3: Analysis of variance for the four reading tasks.
Source SS df MS F
Within 158065.12 20 7903.25
Between 135687.40 3 45229.12 13.96*
Total 194367.50 60 3239.45

*p<.000.

Discussion

There was a significant difference between the reading rates for oral reading and for three silent reading tasks; the subjects' performance dropped by an average of one-third when the subjects moved from the oral reading task to the silent reading and studying tasks. This drop in performance can be accounted for by the change in the demands of the task. In the oral reading task, the subjects could simply focus on reading, but in the silent reading and studying tasks, they were told that they would need to answer specific questions about the material they had just read. Although the purpose of the questions on the silent reading task was simply to check for comprehension, the ultimate effect was to create two studying tasks. When the subjects discovered that the questions were highly specific, their reading rates dropped substantially; as a result, there was no significant difference between their rates on the silent reading and the studying tasks.

Table 4: T-scores and probability for post hoc comparisons of the four reading tasks.
Silent Reading Studying Scanning
Oral reading t=6.47 t=3.26 t=-2.84
p=.000 p=.004 p=.009
Silent reading t=.018 t=4.10
p=.934 p=.001
Studying t=4.38
p=.001

The situation for scanning was different. Since the subjects were asked the question before they began to read, they knew the exact question they were trying to answer. As a result, they were able to disregard information that was irrelevant to the question and to process it only at the recognition level; they did not need to retain information for future reference. On this task, there were significant differences, depending on the scanning strategies of the readers. Those who were familiar with the layout of recipes had a decided advantage on this task because they knew approximately where to look for the information.

The results of this investigation clearly demonstrate the influence of the reading task on the resultant reading rate. Simple oral reading, with no further task demands, resulted in an average reading rate of 135.9 wpm, whereas both the silent reading task and the studying task resulted in significantly lower average scores: 104.7 wpm and 105.8 wpm respectively. As implemented in this study, the silent reading and studying tasks required additional cognitive processes, such as memorization, which are not usually part of the reading process. With regard to scanning, a task in which the subjects knew the information they were looking for in advance, the average rate increased dramatically. This change may be due, in part, to the fact that the reading was done silently, and thus required no vocalizations that would slow down the reading process or add an extra cognitive burden. It could also be explained as a perceptual phenomenon that uses units of spatial frequency that are larger than the single braille cell.

A comparison of the results of this investigation with the taxonomy described by Carver (1981, 1990) clarifies why a task-analysis approach to the reading process is so useful. The demands of a reading task determine the approach the reader uses to complete that task. Thus, when one is looking for a single piece of information, recognition is the key feature. When one is studying for a test, however, other cognitive strategies come into play: rehearsal and chunking. When one engages in silent reading for entertainment, one needs to keep enough information in STM to follow the story, which requires slightly more attention than does recognition. This is the task that Carver referred to as rauding.

In conclusion, the reading rates of expert adult readers, determined in this study, varied greatly. Many of the subjects read at rates that were significantly faster than the 90 wpm often reported in the literature on braille reading. More important than any absolute reading rate in wpms, however, is the fact that reading rates are controlled by the cognitive tasks required or the reader and the purpose of the reading task itself; for example, oral reading is usually not done for the benefit of the reader. Hence, any measure of the reading rate for braille reading must take into consideration more than a perceptual process of word recognition.

This initial investigation should set the stage for a series of studies that can encompass the whole realm of braille reading, including research on the investigations of perception of higher-order units (words), the effect of the structure and complexity of texts on reading rates, comparative studies of reading similar material in print and braille, and comprehension in oral braille reading. Such studies would provide information on the order and sequencing of braille reading instruction and realistic necessary accommodations for individual braille readers.

References

Adams, M.J. (1990 ). Beginning to read and learning about print. Cambridge: MIT Press.

Birns, S. (1976). Review of literature on braille reading. New Outlook for the Blind, 70, 392-397.

Burklen, K. (1932). Touch reading of the blind (F.K. Merry, Trans.). New York: American Foundation for the Blind.

Caton, H. & Bradley, E.J. (1978-79, Fall). A new approach to beginning braille reading. Education of the Visually Handicapped, 66-71.

Carver, R.P. (1981). Reading comprehension and reading theory. Springfield, IL: Charles C Thomas.

Carver, R.P. (1990). Reading rate: A review of research and theory. New York: Academic Press.

CEARSVH. (1991). Literacy for blind and visually impaired students. RE:view, 23, 159-162.

Crandell, J.M. & Wallace, D.H. (1974). Speed reading in braille: An empirical study. New Outlook for the Blind, 68, 13-19.

Flanigan, P.J. & Joslin, M.S. (1969). Patterns of response in the perception of braille configurations. New Outlook for the Blind, 63, 232-244.

Foulke, E. (1964). The comprehension of rapid speech by the blind-Part 2. Louisville, KY: Nonvisual Perceptual Systems Laboratory, University of Louisville.

Foulke, E. (1991). Braille. In M.A. Heller &W. Schiff eds.), Psychology of touch (pp. 219-233). Hillsdale, NJ: Lawrence Erlbaum Associates.

Freeman, I.M. & Dunden, W.J. (1990). Physics made simple. New York: Doubleday.

Grunwald, A.P. (1966). A braille-reading machine. Science, 154, 144-146.

Lorimer, J. & Tobin, M.J. (1979). Experiments with grade 2 braille codes to determine their effect on reading speed. Journal of Visual Impairment & Blindness, 73, 324-328.

Lowenfeld, B., Abel, G.L., & Hatlen, S.C. (1969). Blind children learn to read. Springfield, IL: Charles C Thomas.

Mangold, S.S. (1978). Tactile perception and braille letter recognition: Effects of developmental training. Journal of Visual Impairment & Blindness, 72, 259-266.

Martin, C.J. & Sheffield, C. (1976). The effect of telegraphic prose on the reading behavior of blind and sighted students. Journal of Applied Psychology, 61, 513-519.

McBride, V. (1974). Explanation in rapid reading in braille. New Outlook for the Blind, 68, 8-12.

Meyers, E., Ethington, D., & Ashcroft, S. (1958). Readability of braille as a function of three spacing variables. Journal of Applied Psychology, 42, 183-165.

Millar, S. (1975). Effects of tactual and phonological similarity on the recall of braille letters by blind children. British Journal of Psychology, 66, 193-201.

Millar, S. (1977). Tactual and name matching by blind children. British Journal of Psychology, 68, 377-387.

Newman, S.E. , Hall, A.D., Foster, D.J. & Gupta, V. (1984). Learning as a function of haptic discriminability among items. American Journal of Psychology, 71, 359-372.

Newman, S.E., Hall, A.D., Ramseur, D.J., Foster, D.J., Goldston, D.B., Decamp, B.L., Granberry-Hager, S.P., Lockhart, J.L., Sawyer, W.L., & White, J.E. (1982). Factors affecting the learning of braille. Journal of Visual Impairment & Blindness, 76, 59-64.

Nolan, C.Y. & Kederis, C.J. (1969). Perceptual factors in braille word recognition. New York: American Foundation for the Blind.

Olson, M.R. (1976). Faster braille reading: Preparation at the reading readiness level. New Outlook for the Blind, 70, 341-343.

Olson, M.R., Harlow, S.D., & Williams, J.D. (1975). Rapid reading in braille and large print: An examination of McBride's procedures. New Outlook for the Blind, 69, 392-395.

Olson, M.R., Harlow, S.D., & Williams, J.D. (I 977). An evaluation of McBride's approach to rapid reading for braille and large print readers. Education of the Visually Handicapped, 9, 16-23.

Pick, A.D., Thomas, M.L., & Pick, H.L. (1966). The role of grapheme-phoneme correspondences in the perception of braille. Journal of Verbal Learning and Verbal Behavior, 8, 298-300.

Rex, E., Sowell, V, Koenig, A., Sveen, E.M., & Cheadle, B. (1991). Comments on the position paper on literacy of the Council of Executives of American Residential Schools for the Visually Handicapped. Re:view, 23, 67-74.

Rosen, M. (1991, August). The empress of the sea. Ellery Queen Mystery Magazine, pp. 37-43.

Smith, F. (1988). Understand reading. Hillsdale, NJ: Lawrence Erlbaum Associates.

Trimble, S. (1991). In the shadow of the city: A history of the Loring Park neighborhood. Minneapolis: Minneapolis Community College Foundation.

Uniform Type Committee. (1908). The report of the Uniform Type Committee. Outlook for the Blind, 1, 154-162.

Uniform Type Committee. (1910). The report of the Uniform Type Committee. Outlook for the Blind, 4, 73-81.

Uniform Type Committee. (1912). The report of the Uniform Type Committee. Outlook for the Blind, 5, 1-12.

Uniform Type Committee. (1913). The report of the Uniform Type Committee. Outlook for the Blind, 7, 1-48.

Uniform Type Committee. (1915). The report of the Uniform Type Committee. Outlook for the Blind, 9, 1-92.

Wieten, W. (1989). Psychology: Themes and variations. Belmont, CA: Wadsworth.

Wormsley, DR. (1981). Hand movement training in braille reading. Journal of Visual Impairment & Blindness, 75, 3 27-3 3 1.

Marie Knowlton, Ph.D., associate professor, Department of Educational Psychology, University of Minnesota, 178 Pillsbury Drive, S.E., Minneapolis, MN 55455; E-mail: Know1001@maroon.tc.umn.edu. Robin Wetzel, Ph.D., faculty, Social Science Division, Minneapolis Community College, 1501 Hennepin Avenue, Minneapolis, MN 55403.