Psychological Reports, 1969,24,671-674. @ Psychological Reports 1969

A MULTI-SOLUTION ANAGRAM TASK

ROBERT A. HICKS, MARALEE J. HICK, and HAROLD MANSFIELD

San Jose State College, Cupertino School District, and Regis College

Summary.

A series of multi-solution anagrams, designed to improve on previously used anagram tasks, were standardized on 870 college students. A table of solution-word frequencies is presented. Correlational data pertinent to the use of these anagrams are also given.

A few studies have utilized multi-solution anagram tasks to measure aspects of problem solving, e.g., Maltzman and Morrisett ( 1952, 1953a, 195 3b ) , Ammons and Ammons (1959), Hicks (1964), and Mayzner and Tresselt (1966).

Mayzner and Tresselt (1958) demonstrated the pronounced effect that solution-word frequencies had on the solution times of anagrams. Recently, Mayzner and Tresselt ( 1966) pointed out that the Maltzman and Morrisett data (1952, 1953a) were confounded because they failed to consider the word frequencies of their anagram solutions. Mayzner and Tresselt ( 1966) felt that they had controlled for this factor with solution-word frequencies taken from the Thorndike-Lorge ( 1944) count. In an attempt to measure the novelty of anagram solutions, Hicks ( 1964) used the Thorndike-Lorge word count to establish the frequency of solution-words to a set of multi-solution anagrams.

Hicks' procedure was criticized because in several instances the Thorndike-Lorge frequencies did not correspond to obtained solution-word frequencies. This paper presents a multi-solution anagram task together with frequency and other data appropriate to the evaluation of performance on this task.

METHOD AND RESULTS

Ss were 870 undergraduate college students whose responses provided the basis for the solution-word frequencies. A random sample of 50 male and 50 female Ss was drawn from this group for the purpose of deriving the correlational data.

The 34 anagrams each composed of 5 letters, have with one exception (STOAC) , at least 4 possible word solutions. In constructing these it was thought that ideally each anagram should have several solutions and each solution should represent a point along a frequency of usage continuum.

By inspecting the Thorndike-Lorge word count, it was determined that more 5-letter combinations met these criteria. Then anagrams were developed by systematically picking from the Thorndike-Lorge count 5-letter combinations that formed more than 2 words. From these a final list, which best fit the aforementioned criteria, was selected.

The anagrams were printed in lower case letters (from pilot data, lower case letters were perceived with greater accuracy) , in booklets with instructions on the first sheet, so that the task was relatively self-administering. Essentially, Ss were instructed that ( 1 ) the anagrams all had more than one solution, ( 2) they should try to give as many solutions as possible, and ( 3) Ss were given 2 multi-solution practice problems after which any procedural questions were answered. A time limit of 20 min. was imposed. The order of the anagrams was inverted in every other booklet in an attempt to cancel any such effects in the over-all analysis of the data.

TABLE I
ANAGRAMS AND SOLUTION* FREQUENCIES GIVEN By 870 COLLEGE STUDENTS

Anagram Solutions and Their Frequencies
apres pears 319 rapes 312 reaps 250 spear 230 spare 212 pares 159 apres 30
arcte crate 227 carte 199 trace 172 react 128 cater 45 caret 23
selat tales 345 slate 270 stale 125 steal 107 least 104 teals 37
netso tones 293 stone 190 notes 143 steno 60 onset 47 seton 1
baset beast 119 beats 116 bates 97 baste 96 abets 12 tabes 11
palse lapse 390 peals 128 leaps 101 pales 92 sepal 68 pleas 30 spale 2
ospre spore 280 pores 254 ropes 208 prose 193 poser 23
maste steam 327 teams 285 meats 243 mates 242 tames 168
idset tides 280 diets 117 edits 31 sited 9 deist 1
treim merit 163 timer 155 remit 105 mitre 50 miter 14
aslev slave 387 vales 164 salve 100 veals 84 laves 67 valse 37
arens snare 209 nears 63 earns 46 nares 46 saner 3
tesim times 287 mites 228 smite 196 emits 33 items 26
sneir reins 278 siren 103 resin 40 risen 37 rines 34 rinse 26
ursec curse 328 cures I11 cruse 7 sucre 3
prasc scrap 326 craps 188 carps 63 scarp 5
stoac coats 463 coast 99 ascot 7
narog groan 223 organ 46 argon 27
nalep plane 251 panel 173 penal 95
asett taste 254 state 177 teats 47 tates 4 testa 1
rhate heart 108 hater 101 earth 17 rathe 9
epasc space 240 paces 192 capes 187 scape 162
telap plate 238 petal 189 pleat 50 leapt 39
miesl miles 301 smile 286 limes 237 slime 231
sitel tiles 243 stile 140 lites 54 slite 9 islet 1
crase scare 219 races 174 cares 132 acres 29 caser 8
orecd cored 196 decor 192 credo 49 coder 10
selah hales 163 shale 152 heals 135 leash 82
arest stare 274 tears 179 rates 120 stear 82 tares 66 aster 3
snite tines 141 stein 89 nites 86 inset 12
rehas hares 188 share 186 hears 164 shear 130
evres sever 279 serve 226 verse 105 veers 87
rilef rifle 318 flier 80 filer 75 lifer 15
kaset takes 213 skate 182 stake 123 steak 120 teaks 32
*Word-solutions were defined by The Oxford Universal Dictionary.

A MULTI-SOLUTION ANAGRAM TASK

The anagrams were administered to groups during the regular class period. When the data had been collected, the solution-word frequencies were tabulated. Then the sample of 100 Ss was selected and these Ss' anagram responses, i.e., number correct and number of errors, were correlated with a number of available measures, GP A, College Entrance Examination Board ( CEEB) Verbal and Math scores, and sex, using the F ACV AR program at the University of Denver Computer Center.

The anagrams and the frequency counts obtained for each solution-word are presented in Table 1. Table 2 gives the Pearsonian correlations among the anagram variables, and sex, and measures of Ss' abilities. The anagrams presented in Table I were designed primarily for use in problem-solving research.

This set of anagrams is an attempt to overcome certain difficulties, cited earlier in this paper, associated with multi-solution anagrams by providing empirically determined solution-word frequencies. It is thought that the frequency data given in Table I could be used to measure the novelty of anagram solutions, as have Ammons and his colleagues ( e.g., Robertson & Ammons, 1962) .

TABLE 2

CORRELATIONS BETWEEN NUMBER OF CORRECT RESPONSES AND ERRORS AND GPA, CEEB MATH AND VERBAL SCORES, AND SEX

Variable Anagram Variables

No. Correct Errors

CEEB Math .29* .09

Verbal .18 -.08

GPA .10 -.08

Sex .07 .11

Anagram Errors .16

*Significantly different from Rho (p < .01).

The data given in Table 2 indicate that the correlations between anagram performance and the subject-variables measured are uniformly low; with one exception, none are significantly different from zero. The correlational data seem at least partially to negate these subject variables as potential sources of confounding.

REFERENCES

AMMONS, R. B., & AMMONS, C. H. A standard anagram task. Psychological Reports, 1959, 5, 654-656.

HICKS, R. A. The effects of induced muscular tension on the performance of complex tasks. Unpublished doctoral dissertation, Univer. of Denver, 1964.

MALTZMAN, I., & MORRISETT, L., JR. Different strengths of set in the solution of anagrams. Journal of Experimental Psychology, 1952, 44, 242-246.

MALTZMAN, I., & MORRISETT, L., JR. The effects of single and compound classes of anagrams on set solutions. Journal of Experimental Psychology, 1953, 45, 345-350. (a)

MALTZMAN, I., & MORRISETT, L., JR. Effects of task instructions on solution of different classes of anagrams. Journal of Experimental Psychology, 1953, 45, 351-354. (b)

MAYZNER, M. S., & TRESSELT, M. E. Anagram solution times: a function of letter order and word frequency. Journal of Experimental Psychology, 1958, 56, 376-379.

MAYZNER, M. S., & TRESSELT, M. E. Anagram solution times: a function of multiple- solution anagrams. Journal of Experimental Psychology, 1966, 71, 66-73.

ROBERTSON, D. G., & AMMONS, R. B. "Problem" norms for the Standard Anagram Task. Proc. Mont. A cad. Sci., 1962, 21, 97-104.

THORNDIKE, E. L., & LORGE, F. The teacher's word book of 30,000 words. New York: Teachers Col., Columbia Univer., Bur. of Publ., 1944.

Accepted March 15, 1969.