Solving anagrams as a function of word frequency, imagery and distribution of practice.

Abstract

While previous research has consistently demonstrated that anagrams with solutionwords of high frequency are easier to solve than anagrams with solution-words of low frequency, the results for solution-word imagery have been contradictory. Experiment I involved the use of four six-item lists which varied factorially on imagery and frequency and were equated for meaningfulness and total bigram-position frequency. Eighteen Ss attempted to solve these anagrams under conditions of distributed practice and 18 Ss under massed practice. Data for solution latency and number of solutions revealed that anagrams were easier to solve if the solution-words were of high imagery or high frequency (ps < .001). Distribution of practice and practice over 24 problems did not influence performance. Experiment n replicated the effects of solution-word imagery and frequency with a different sample of words and with more precise controls (ps < .025). The results were discussed in terms of the processes involved in the storage and retrieval of words from semantic memory and die role that retrieval processes play in solving anagrams.SOLVING ANAGRAMS involves alternating between letter rearrangement, word retrieval, and word-anagram comparison activities (see Johnson, 1972). The major purpose of the present studies was to determine the effects of solution-word frequency and imagery on the ease of retrieving words to solve anagrams.Anagrams and thinking-of-a-word problems are more easily solved if the solutions are high-frequency words than if the solutions are low-frequency words (e.g., Mayzner & Tresselt, 1958;Stratton & Wathen, 1972). This effect is usually attributed to the tendency to retrieve words from memory in an order determined by frequency of experience, with high-frequency words being retrieved before low-frequency words, the spew law of Underwood and Schulz (1960). Also words of very low frequency may never have been stored in memory (Duncan, 1973).The verbal learning literature suggests that word imagery may also influence performance on anagram problems. Several studies nave shown•This is an expanded version of a paper presented at the 1973 meetings of the Midwestern Psychological Association in Chicago. Requests for reprints should be addressed to R.

Effects of Solution Word Attributes on Anagram Difficulty: A Regression Analysis

“…Dewing and Hetherington (1974) found that high imagery solutions significantly facilitated anagram solving when solution word frequency, bigram frequency and associative meaningfulness (Noble, 1952) were controlled. This result was replicated with different materials and extra controls for bigram frequency and versatility (Solso et al 1973) factors by Stratton, Jacobus and Leonard (1975). The simplest conclusion seemed to be that words that have high imagery values are more available than words of low imagery value.…”

“…In view of previous reports of significant imagery effects in anagram solving (Dewing and Hetherington, 1974;Stratton, Jacobus and Leonard, 1975), it is noteworthy that neither imagery nor its close associate, concreteness, had any discernible effect on solution scores in the present study. These other experiments manipulated imagery by selecting words from the extremes of the scale while the present study drew a random sample from the entire range.…”

“…So, multiple regression studies do not suffer from those shortcomings of factorial experiments on word attributes that were pointed out in Clark's (1973) well known paper on the "language-as-a-furedeffect" fallacy. The previous studies that reported imagery effects (Dewing and Hetherington, 1974; Stratton, Jacobus and Leonard, 1975) are open to Clark's critique, in that they did not demonstrate that the effect was reliable over words. Despite some recent dissent (Wike and Church, 1976), Clark's critique is widely accepted and the imagery effect on anagrams would have a more certain status if it had been demonstrated with words treated as random rather than fixed effects.…”

“…The word rating studies of Paivio et al (1968), Stratton, Jacobus and Brinley (1975) and Gilhooly and Hay (1977) indicate that concreteness, associative meaningfulness, familiarity, objective frequency and age-of-acquisition are significantly correlated with word imagery (and with each other). Of these variables, meaningfulness and frequency were controlled by Dewing and Hetherington (1974) and Stratton, Jacobus and Leonard (1975), but the remainder were not.…”