Judging age from handwriting done with and without visual feedback

Bulletin of the Psychonomic Society 1993.31 (2), 111-113 Judging age from handwriting done with and without visual feedback EUGENE A. LOVELACE, BETH A. VELLA , and DONNA M. ANDERSON Alfred University, Alfred, New York In a prior study (Lovelace & Aikens, 1990), young and old adults wrote 10 common words with their eyes open and again with their eyes closed. The present study used writing samples from 20 old and 20 young writers (67-85 years vs. 18-24 years) in the prior study. Thirty college students judged (1) which were written with eyes closed (paired comparisons, same word written by same person) and (2) whether the writer was young or old. Both tasks were difficult; performance was significantly above chance, but not good (71% correct on open/closed, 65% on young/old, both ps < .001). Accuracy of open/closed decisions was no higher when judging old writers than young writers, implying that the motor skill of handwriting is well preserved in aged individuals; they do not have an increased need of visual feedback to write. Judges also reported cues they had used to make these decisions. Accuracy of age decisions was significantly related to the reported use of style cues (73%correct for subjects reporting they used that cue vs. 53% when not reported, chance = 50%). The questions of interest in this study were the following: How well can untrained judges tell the age of a person from a sample of handwriting? Can they judge this more accurately if the person wrote with his/her eyes closed? What sorts of cues might be used to make this judgment? To provide an initial assessment of answers to these questions, we made use of handwriting samples gathered in another study (Lovelace & Aikens, 1990). In that study, young and old adults wrote 10 common words of five or six letters (no is or ts) with their eyes open and again with their eyes closed. The present study used writing samples from 20 young adults (18-24 years old) and 20 old adults (67-85 years old; M = 73.6, SD = 5.2) in the prior study. Preliminary analyses of these writing samples, plus the lack of any appreciable age effects on other kinesthetic/motor tasks in the Lovelace and Aikens study, indicated that the motor task of writing might be well preserved in the elderly. The present study allowed a test of discrirninability as to whether words were written with eyes open or closed when the same words, written in these two ways by the same individual, were simultaneously present for direct comparison. If the kinesthesis involved in writing, or the accuracy of compiling and executing the motor program to write the words, showed substantial age effects, then the aged would be expected to suffer more when deprived of the opportunity to visually control their writing. Correspondence should be addressed to E. A . Lovelace, Psychology Division , Alfred University , Alfred, NY 14802 . METHOD Subjects Thirty students enrolled in introductory psychology courses at Alfred University participated as an option for extra course credit. They ranged in age from 18 to 23 years (M = 19.4, SD = 1.3) . Procedure Words written with eyes closed are often at a slant on the page . To eliminate this cue, words were cut from the page . For each writer, the two versions of each word (written with eyes open vs. closed) were placed side by side in two columns; the left-right position of the eyes-closed vers ion was random, with the restriction that this version occur equally often in the two columns . The 10 words appeared in the same order in 81140samples. Order of the samples with respect to age was random . The students were asked to make two judgments: (l) for each pair of words, they were to decide which was written with eyes closed ; and (2) for the entire set of 10 word pairs , they were to decide whether the writer was young or old . After completion of all judgments, the students were asked to describe any cues in the handwriting they were aware of using when making each of the two types of decisions . RESULTS OveraU Levels of Judgment Accuracy Both tasks were viewed by the judges as difficult, but performance was above chance . The open/closed decisions were correct 71 % of the time (SD = 6%, range 52 %-82 %), which differs significantly from chance (50%; t = 20.5, p < .(01). Contrary to what one would expect if the aged have reduced kinesthetic-sense ability and so must rely more on visual guidance in writing, there was no difference in the success of these judgments for young (71 %) versus old (72%) writers (t < 1.0) . The young/old choices yielded 65 % correct decisions (SD = 15%, range 33%-90%). Although accuracy of this judgment varied widely, the mean percent correct was 111 Copyright 1993 Psychonomic Society, Inc . 112 LOVELACE, VELLA, AND ANDERSON greater than chance (I = 5.3 , p < .(0 1). For these decisions, the judges as a group operated in an unbiased manner; half the samples were from young writers, and the mean (and median) proportion " young" responses was .50 (range .38-.65) . Reported Cue Use and Performance The cues reported by the students were assigned to categories of similar responses . For the open/closed task, the mean number of cues reported was 3.2 (range 0-6). They fell into the following 10 categories (with ns) : spacing (17), neatness (14), size (13), flow/connected (11), letters closed (9), slant (9), spelling (8), uppercase/lowercase (6), pressure (4), and shakiness (4). (Note that since we had cut the words from the page, any remaining slant on the page was largely an artifact of that procedure; thus, no analysis was performed on the slant cue .) For each of these cues, the data were partitioned on the basis of reported use (or not) of that cue. None ofthe reported cues was related to performance. The proportion of correct decisions , as a function of reported use versus nonuse of each cue, never differed by more than .05, and in three cases it actually showed a reversal of direction . None of these differences approached statistical significance . Clearly, there was no substantial relationship between the use of any of the reported cues and accuracy of open/closed decisions. For the young/old decisions, the mean number of reported cues was 1.7 (range 0-4). They fell into the following nine categories with (ns) : style (18), neatness (6), size (6), similarity of the open and closed versions (5), spelling (4), shakiness (3), slant (3), spacing (3), and cursive/print (3). Only one cue appeared to be clearly related to performance. The cue that was most widely re- Table 1 Samples of Reports Categorized as Containing the Style Cue • • • • • • • • • • • • • • Old have letters more evenly distributed; young put more emphasis on first and last letters. Old have precise lettering with sort of authority; young didn 't have precise lettering. Young use " bubble-like" letters, mix in cursive and print ; old write more correctly as cursive was taught. Old have a fancier style of handwriting; youn (...truncated)