Twelve (6 men, 6 women) young (mean age 22.8 years, range 20 - 28), and twelve (6 men, 6 women) older community-resident volunteers (mean age 68.9 years, range 61 - 75) who were mobile and generally in good self-reported health, participated in this study.

Participants were tested for near and far acuity, contrast sensitivity, and astigmatism. They were also surveyed about their visual limitations and problems. Although the vision of adult observers can usually be improved refractively (Owsley, Sekuler & Siemsen, 1983), to enhance the real world generalizability of the results, and considering that all participants had reasonably good acuity (i.e., none worse than 1.75 minarc), all testing was carried out using participantsí presenting optical correction.

Two younger participants and four elderly participants presented with an astigmatism. Self-reported visual problems were also more common among the elderly participants. One young participant reported a high level of light sensitivity; one older person reported a floater, two reported blurred vision when tired, and one experienced occasional double vision. No one reported having any chronic visual diseases (e.g., glaucoma, cataracts). Two young participants reported general health problems: one suffered from occasional headaches, and the other had asthma. Five of the elderly participants complained of health problems: two suffered from high blood pressure, two had high cholesterol, and one had osteoarthritis. The native reading language for all participants was English, although one young participant reported that English was his second language. Elderly participants were more likely than younger participants to require glasses (especially to read) and on average, more time had elapsed since they had received their last prescription (see Table 1). There were no age differences on either years of education (t(22) = 0.53 p >.05;) or WAIS-R vocabulary scores (t(22) = 0.70 p >.05).

Table 1

Background Information for 12 Young and 12 Elderly Participants
 

	Young	Elderly
English first language	11	12
Use corrective lenses	5	12
If use corrective lenses
Years since last prescription	0.80	3.25
Single-vision glasses	4	5
Contact lenses	3	1
Multifocal lenses	0	7
lenses to read	0	10
lenses to drive	5	7

For inclusion of their data in the study, participants were required to answer correctly at least one of the two comprehension questions associated with each passage, and overall to answer correctly 13 of the 16 questions. A liberal criterion was used to determine correct answers and thus the reading of passages was indoubt for those failing this criterion. For this reason the data from seven participants (five elderly males, an elderly female, and a young female) were not included in the study. Analysis of variance revealed that the elderly participants eliminated from the study did not differ significantly from those used in the study on any participant measure nor dependent measure (ps ranging from .973 to .164) except near peak CS frequency (M = 2.33 for those eliminated and M = 3.67 for those used in the study t(16) = 2.28, p < .05), and a mean vocabulary score slightly (but not significantly lower). An examination of the mean scores of the younger participant replaced indicated a slightly higher contrast sensitivity at lower frequencies for both near and far test distances and a lower acuity for the far test distance. No other scores fell outside the range of scores obtained by the other young participants. The data were replaced by testing seven additional participants from the appropriate age/sex groups and test order.

Vision tests: Near (36 cm) and far (610 cm) acuity were assessed, with their presenting correction, using custom Landolt-C Acuity Charts. Near contrast sensitivity was determined for gratings of 1, 2, 4, 8 and 12 c/deg, and far contrast sensitivity was measured for gratings of 3, 6, 12, 24 and 36 c/deg utilizing the Vision Contrast Test System (VCTS) charts for near and far distances respectively. Participants were screened for astigmatism utilizing the AO Scientific Instruments Nearpoint Rotochart, and Greenís Astigmatic Test Card for near and far distances, respectively.

Self-report survey: Background demographic information was obtained via. oral questions (see Appendix A). Information on visual problems and reading was obtained using a participant completed questionnaire (see Appendix B).

Print materials: Eight fonts were evaluated in the study (see Appendix C); four were serif fonts (Century Old Style, Times New Roman, Garamond Condensed and Sabon), and four were sans serif fonts (Times New Roman Sans, Univers Condensed, Bell Gothic and The Sans). A serif font (Palatino) was used to familiarize participants with the legibility and readability tasks. It was also the font in which all non-test stimulus (e.g., instructions) were printed. The print materials were printed on plain white paper on a Laser Writer 16/600 PS printer at 600 dots/inch and presented at a luminance of 100 cd/m2 as measured using a Minolta LS110 spot photometer.

The legibility of each font was evaluated using charts composed of lowercase letters in each of the nine fonts. Each chart contained 21 rows each of 16 letters arranged in four, four-letter horizontal clusters. The critical element of the print (i.e., 1/5 x-height) ranged in size from 0.5 minarc (equivalent to a 6/3 target) for the smallest size up to 2.5 minarc (6/15) in .1 minarc steps. Each line was composed to balance letters that were easy (e.g., p, z, c, m) or difficult to discriminate (e.g., e, i, j, t). All lines of the same letter size on each font chart contained the same letters (see Appendix D), but arranged in a different randomized order. Each four-letter cluster contained at least one riser or descender. Only one riser and/or one descender was permitted in each letter cluster. The position of ascender and descender letters was balanced across each of the four positions in the clusters within a chart. To reduce possible differential "top-down" effects, the clusters did not form words or "word-like" groups. Uniform inter-cluster and row spacing was maintained across lines of the same size for each chart (see Table 2). To provide uniform open spacing between clusters a blank space in non-proportional font (e.g., Courier) was used.

Table 2

Uniform Inter-cluster Spacing and Row Height for Nine Legibility Charts
 

Letter Size (minarc)	Vertical Spacing (points)	Horizontal Spacing (points)	Letter Size (minarc) ctd.	Vertical Spacing. (points) ctd.	Horizontal Spacing. (points) ctd.
2.5	228	139	1.4	129	78
2.4	220	133	1.3	120	73
2.3	211	128	1.2	111	67
2.2	204	123	1.1	100	61
2.1	192	116	1.0	91	56
2.0	185	111	0.9	82	50
1.9	173	106	0.8	72	44
1.8	167	100	0.7	65	39
1.7	156	94	0.6	55	34
1.6	146	88	0.5	46	28
1.5	137	84

For each of the nine fonts, passages of 200 words were constructed of near equivalent cognitive reading difficulty (standard reading difficulty or grade 7 - 8 reading level) as assessed by the Flesch Reading Ease (M = 66.4, SD = 3.8, range 59.5 to 71.9, possible range 0 - 100, Flesch, 1948), and (grade 9 reading level) Coleman-Liau Grade Level (M = 9.4, SD = 0.9, range 7.8 to 10.6, possible range 1 - 16, Coleman & Liau, 1975). The passages and their instructions were printed at 0.045 inch x-height with equal inter-line spacing (see Appendix E). Two general comprehension questions were included for each passage to ensure the reading of each passage.

Reading ease/discomfort/clarity of type: Each passage was rated on a 7 point scale for Reading Ease, Reading Discomfort, and the Clarity of Type (see Appendix F). These questions were printed in the Palatino font.

Simplicity/complexity: A single short statement printed in each font was used as the basis for rating its Simplicity/Complexity. Font size was the same as that used in the readability passages (i.e., x-height = 0.045 inch). A different random order was used for each participant in each age by sex group (see Appendix G).

Preference: Font preference was assessed using the same sentence as those used to evaluate simplicity/complexity, but with a different random order for each participant in each age by sex group (see Appendix H).

Vocabulary: Vocabulary was assessed using the vocabulary subtest of the Wechsler Adult Intelligence Scale-Revised (WAIS-R; Wechsler, 1981).

Participants were tested in a single self-paced session lasting 1 to 2 hours. Informed consent, and background information regarding issues such as education, health, visual problems were obtained at the beginning of the session. The far and near spatial measures were evaluated at the beginning of the legibility and readability sections, respectively. The repeated measure design was used to prevent differential motivational factors affecting different participants (and thus fonts). In this way motivational factors were controlled by having people act as their own controls and the counterbalancing of the font orders prevented systematic motivational differences in performance for any font.

Visual measures: Prior to the assessment of legibility or readability, visual performance was assessed. Acuity was measured by the determination of the smallest gap in a line of Landolt-Cs that could be correctly identified. This was followed by the identification of contrast sensitivity (the least contrast at which the orientation of lines could be identified) for various increasing frequencies. This information was used to create a contrast sensitivity function (an assessment of sensitivity for each frequency tested), and several composite measures of contrast sensitivity (frequency at which greatest sensitivity occurred, Peak and Mean Sensitivity, and High Cutoff Frequency (the frequency above the Peak Frequency at which sensitivity drops to half Peak Sensitivity). Astigmatism was also assessed.

Legibility: Legibility thresholds, defined as the smallest print line (1/5 x-height in minarc) at which no more than 2 of the letters on each line of 16 were incorrectly identified, were determined for each of the eight fonts. This criterion was used to allow for some confusion between letters while still requiring correct identification. Beginning with a line that the participants indicated they could easily read, observers progressed down the chart, reading each line of successively smaller print until the first line was reached at which more than two errors were committed. Before beginning, observers received a practice trial using the Palatino font to familiarize them with the legibility task prior to testing.

Readability: Two measures of readability, Comfortable Reading Distance and Reading Time, were assessed for each of the 8 fonts. Comfortable Reading Distance between the eye and the plane at which the observer placed the materials was measured using a string as each participant read the task instructions printed in the same font/size as the test passage. Reading time was measured as the time the participants took to read each passage at the instructed "normal comfortable" reading speed. Participants were informed that two general comprehension questions would be asked immediately upon completion of the passage. After both questions were answered, the observer rated the font used in the passage in regard to its "Reading Ease", its "Reading Discomfort" and its "Clarity of Print" on separate 7-point scales (see Appendix F). At the beginning of the readability trials, a practice trial was provided using the Palatino font to familiarize participants with the task.

Simplicity/complexity/preference: Once the readability assessments for all fonts were completed, participants rated the complexity of the letters composing each font on a 7-point scale (see Appendix G). Finally, corresponding preference for the eight fonts was assessed (see Appendix H) using a ranked forced distribution (i.e., 8-point scale using all 8 numbers).

Overall research design: One participant from each Age x Sex group received the exact same stimulus order (legibility, followed by readability, complexity rating, and preference), which required 6 different sets of stimulus orders. Half of the participants in each Age X Sex group received the legibility task followed by the readability/simplicity/ complexity/preference tasks, and the other half received the reverse order (i.e., readability, complexity rating, preference followed by legibility). At the end of the session, participants completed the WAIS-R vocabulary subtest.