The Influence of Familiarity on Word Length Effects through Verbal Recall
Hillary Coleman and Faith Schombs
In the early days of research on short-term memory, Miller (1956) suggested that the magical number 7 ± 2 could be used to account for the span of a person’s attention and the span of immediate memory. The idea of this theory implies that when a person hears a list of 7 ± 2 items, such as numbers, they are usually able to recall the list correctly, but when a greater number of items are asked to be remembered, there may be less accuracy in recall. While this hypothesis may work with some objects and aspects of memory, it has not shown to work as well when keeping material in one’s articulatory loop, specifically words. Under Miller’s theory, every word, regardless of the length and duration it takes to say, would be considered to be one item of 7 ± 2 that we can hold in our memory span; however, more recent research provides a different theory. Baddeley, Thomson, and Buchanan (1975) suggest an alternative theory that “we can keep about 1.5 to 2.0 seconds worth of material rehearsed in the articulatory loop” (Anderson, 2009, p. 154). This theory stems from the idea that the capacity of short-term memory of speech may be better understood by measuring shorter speech units such as syllables and or how long it takes to say a word. To research this hypothesis, Baddeley et al. (1975) conducted a number of studies that used short, usually one or two syllable words, and long, usually four or five syllable words, to present to participants and observe their recall abilities. The overall results of their eight experiments found that memory span is based on the length of material as opposed to the number of items suggested by Miller (1956), thus introducing the concept “word length effect.”
“The word length effect, the finding that lists of short words are better recalled than lists of long words, has been termed one of the benchmark findings that any theory of immediate memory must account for” (Jalbert, Neath, Bireta, & Suprenant, 2011, p. 338). This definition can be explained through research that suggests that
people have greater accuracy in recall of words that are short compared to long words. Since the Baddeley et al. (1975) study, a majority of research on short-term memory and recall ability does consider the word length effect as an important aspect of the capacity of human memory. Hulme et al. (2006) conducted a study that examined serial recall of short words compared to long words, further emphasizing the importance of the word length effect while also accounting for the importance of positioning of words in memory. Again, they found that there was a significant recall advantage for shorter words over long words in most cases. While a significant amount of research on word length effect has focused on the capacity to recall words spoken aloud, the word length effect does not only pertain to spoken word. In a similarly structured study, Wilson and Emmorey (1998) explored the capacity of working memory with deaf participants focusing on the use of sign language, not spoken word. Wilson and Emmorey (1998) found that the structure of working memory for sign language parallels the structure of working memory for speech; longer signs produced poorer memory performance than did lists of shorter signs. The word length effect plays a crucial role in people’s ability to recall words listed aloud or signed, and will be accounted for in the present study with the idea that short words are more likely to be recalled correctly than long words.
Another interesting topic pertaining to memory span is looking at the way familiarity with words affect a person’s capacity and ability to recall them. One of the main studies that provides extensive conclusions about this field is by Turner, Henry and Smith (2000). In this study, the researchers hypothesized that the reason unfamiliar items could be given meaning, thus increasing the memory span for these items, would be due to either an increase in semantic knowledge or an increase in the familiarity of the phonology. These researchers found that pseudo-words were more memorable when they had become phonologically familiar rather than when they had been given a made up meaning. It has also been shown that the longer in length a pseudo-word is, the longer it takes to recognize it (Richards & Heller, 1976). Additionally the findings of Roche, Tolan, and Tehan (2011) suggest that familiar concrete words are remembered at greater rates than unfamiliar abstract words. Through this collection of studies on the effect of the familiarity of words on short term recall, it is evident that familiar words tend to be remembered with greater rates, most likely because these words are used more often in one’s daily life and are more readily available and accessible in their brain’s memory storage.
This idea of familiarity and unfamiliarity can also be tied to the well-known concept of schemas. A schema is a representational structure that contains information that a person commonly links with a certain object, the idea being that through hearing one word in the schema; activation will spread and make other words in the schema
more readily available for memory recall (Anderson, 2009). Stuart and Hulme (2000) suggest that associative links have an effect on the recall of words during short-term memory performance tasks. These links seem to be influenced by long-term memory since it was found that associations between items in long-term memory, thus more familiar items, cause a greater level of recall of words during short-term memory tasks (Stuart & Hulme, 2000). Priming participants with the familiarization of words or pair associations has also improved item recall in short-term memory tasks due to increasing familiarity and associations prior to the tasks (Saint-Aubin& Poirier, 2005). In their study, Saint-Aubin and Poirier (2005) asked participants to recall lists of words in a similar procedure to that of many word length effect studies; first providing subjects with no familiarization of the words in the study, item familiarization, or pair familiarization. It was found that “both item and pair familiarization improved the item recall of low-frequency items to the same extent, suggesting that increased familiarity can account for the co-occurrence effect” (Saint-Aubin & Poirier, 2005, p. 325). The co-occurrence hypothesis, originally proposed by Deese (1960) “suggests that the retrieval of an item is at least in part a function of the characteristics of other list items” (Saint-Aubin & Poirier, 2005, p. 326). This seems to suggest that by increasing familiarity, whether it is long-term familiarity or familiarity created prior to recall tasks through studying a list of words or pairs, unfamiliar words will not be recalled as easily as familiar words. Similar to the activation of associations between types of words on a list, Kolic-Vehoyec and Arar (2005) found that when asking participants to recall a surname, those who were primed with first names had a more pronounced effect of recalling surnames than those primed with non-name words. This result could be attributed to the activation of a name schema by hearing a first name. The familiarity of a word is a vital factor in the recall of words in short term word memory tasks because familiar words are more likely to be activated than unfamiliar words.
Overall, schemas and associations in long-term memory seem to play an important role when asking people to remember familiar versus unfamiliar words. The benefit of schemas was also found in a study by van Kesteren, Ruiter, Fernandez, and Henson (2012) who compared the rates of memory for schemas and less congruent information. Again, schemas were found to usually be better remembered than less congruent information. Pertaining to cognition and brain make-up it is important to note that the “lateral prefrontal cortex plays a critical role in the monitoring or decision processes required for accurate familiarity-based recognition responses” (Aly, Yonelinas, Kishiyama, & Knight, 2011, p. 297). These researchers found that damage to the lateral prefrontal cortex did not impact the recollection of information but did affect the familiarity aspect of recall. Thus it appears that familiarity and word length effect, further explained by the use of schemas, are both factors that help to explain the memory capacity of the human brain pertaining to a person’s ability to remember some words or objects better than others.
The present study draws from the above ideas of memory for familiar words and the word length effect, which provides that it is easier to recall words that take less time to pronounce (Baddeley et al., 1975; Cowan, Wood, Nugent, & Treisman, 1997). The method employed in the present study is closely aligned with that of Hulme et al. (2006), who presented participants with small lists of words, which they were asked to repeat. The researchers combined this method of Hulme et al. (2006) with that of Roche et al. (2011) where only five words were presented in each list. Furthermore, this research drew on ideas from Baddeley et al. (1975), who, in one of their experiments, measured the accuracy of recall by participants after seeing lists of five long or short country names. It was found that for short names, participants correctly recalled 4.17 out of 5 words and for long names, 2.8 out of 5 (Baddeley et al., 1975). While Baddeley et al. (1975) showed that memory span was more efficient for the names of shorter countries, the current researchers expanded this study by comparing the length of place names and level of familiarity in memory recall.
The present study used places at Connecticut College as long and short familiar words and names of foreign countries that are less well known for
long and short unfamiliar words. As the participant pool was drawn from Connecticut College, researchers postulated that these students would consider words pertaining to places at their school as more familiar than the names of foreign countries such as Malta and Estonia. All words used were spoken using the English pronunciation to prevent possible confusion by language differences in foreign pronunciations of the country names. First, it was hypothesized that shorter words in both unfamiliar and
familiar conditions would be recalled in the correct order with greater accuracy than longer words in the perspective conditions, accounting for the word length effect. It was further hypothesized that familiar words, when compared to those of the same length in the unfamiliar condition, would be recalled in the correct order with greater accuracy than unfamiliar words, thus accounting for the familiarity factor in word recall tasks.
The research design for this study was a mixed design. Participants were randomly assigned to one of two groups, those looking at familiar words and those looking at unfamiliar words. Since the true independent variable was the familiarity of the words given, the researchers used a between subjects design. Within the familiar or unfamiliar group, they then used a repeated measures design where participants viewed both long and short words. A manipulation check question was presented at the end to see if participants understood that part of the study was about verbal recall of long versus short words.
From both the Psychology 101/102 Connecticut College students pool and other students on the Connecticut College campus, 26 college students voluntarily signed up to participate in this experiment; thirteen were randomly assigned to the familiar condition, viewing both short words and long words; ten were randomly assigned to the unfamiliar condition, viewing both short words and long words. Of these 26 participants, 20 were women and 6 were men, with the average age being 20. Based on
class year, the majority of participants were Juniors (n = 13), followed by Seniors (n = 5), Sophomores (n = 4), and Freshmen (n = 4). Eleven participants self-identified their ethnicity as being White, 6 identified as being Asian-American or from an Asian country, 5 identified as being Latino or Hispanic, 3 identified as Black or African-American, and 1 participant did not identify their ethnicity. The study was conducted through 15 minute in person, one-on-one sessions, where participants heard recordings of word lists. The demographic questionnaire at the end was completed by all participants. Participants that were enrolled in psychology courses received credit for their participation.
Besides the Informed Consent (see Appendix A) and the Debriefing Form (see Appendix C), participants also heard a series of word list recordings. There were a total number of 5 variations of said recordings for each condition (short familiar, short unfamiliar, long familiar and long unfamiliar). The recordings were created by using a third party voice to ensure that participants would not link the voice they were hearing in the recordings to that of the researchers. Participants were also asked to fill out a Demographic questionnaire (see Appendix B) that contained basic questions about age, gender, class year, and ethnicity at the end. On this questionnaire, participants further indicated if English was their first language and if they had a learning disability that they thought would affect the results of their participation in this study. There were in total 3 participants that had such a learning disability and their results were discarded from the study.
Participants first filled out an informed consent and then were asked to complete the study one at a time in a quiet room on the second floor of Shain Library of
Connecticut College. In the familiar word condition, a participant heard recordings of five familiar words read at a 1.15 second rate. They had 15 seconds to recall the list of words in the correct order, and were prompted with the word “ready” to indicate that another list of words was going to be read. In total, there were 5 recordings of short familiar words and 5 recordings of long familiar words. The short familiar words were drawn from the list: Cro, Shain, Plex, Bill, Smith, Wright, Plant, and Park. The long familiar words were drawn from this list: Arboretum, Morrison, Camelympics, Graduation, Mohegan, and Hamilton. The unfamiliar word condition was identical to the familiar word condition except for the fact that the participant heard five different, short or long, unfamiliar words. Short unfamiliar words were: Chad, Laos, Chile,
Tonga, Kenya, Burma, and Malta. Long unfamiliar words were: Albania, Somaliland, Cameroon, Estonia, Mongolia, and Turkmenistan. Counterbalancing was used in both
conditions in that the word lists were presented randomly to participants. After completing the word memory recall task, participants were asked to fill out a basic demographic questionnaire that included a manipulation check question. Finally the participants received a debriefing form and research credit if they needed it.
To evaluate the hypotheses that shorter words, regardless of condition, would be recalled in the correct order with greater accuracy than longer words and that familiar words would be recalled in the correct order with greater accuracy than unfamiliar words, a one-way Analysis of Variance (ANOVA) and two Paired Samples t-tests were employed. The hypotheses are contingent on words being recalled in the correct order that they were played in, however many participants recalled some words in the correct order and not others. To account for this discrepancy, 2 sets of tests were completed, the first based on recalling words in the correct order, measured by recording the number of words recalled in the correct order until a mistake was made, and the second accounting for overall number of words recalled. No significant differences by gender or age were found in the results.
The one-way ANOVA for correct order of words comparing participants results from the familiar and unfamiliar conditions showed a significant difference between familiar and unfamiliar word recall for both short words, F(1, 21) = 11.59, p = .003, and long words, F(1, 21) = 16.822, p = .001, indicating that participants in the familiar condition better recalled words of the perspective lengths in the correct order than did participants in the unfamiliar condition. Means and standard deviations of number of words recalled in the correct order for familiar and unfamiliar, and short and long words are presented in Table 1. Paired Samples t-tests that were used to examine word length effect in participants in the familiar condition indicated that participants recalled a greater number of short familiar words (M = 4.60) in the correct order than they did long familiar words (M = 2.88), t(12) = 6.74, p < .001. Paired Samples t-tests also indicated that participants in the unfamiliar condition recalled, in order, a greater number of short familiar words (M = 3.24) than long unfamiliar words (M = 1.24), t (9) = 8.66, p < .001.
Results were also analyzed to account for the total number of words recalled correctly regardless of order. A one-way ANOVA indicated both that participants in the familiar condition recalled a greater number of short words than did participants in the unfamiliar condition, F(1, 21) = 7.13, p= .014) and that participants in the familiar condition remembered more long words than did participants in the unfamiliar condition, F(1, 21) = 17.38, p < .001. Means and standard deviations of number of words recalled correctly for familiar and unfamiliar, and short and long words are presented in Table 2. Paired Samples t-tests indicated that familiar short words (M= 4.80) were recalled with greater accuracy than were familiar long words (M = 4.20), t(12) = 3.79, p = .003. It was also revealed through Paired Samples t-tests that participants recalled a greater number of short unfamiliar words (M = 3.96) than long unfamiliar words (M = 2.90), t(9) = 4.41, p = .002.
Examining the effects of familiarity and the word length effect, this research hypothesized that shorter words in both unfamiliar and familiar conditions would be recalled in the correct order with greater accuracy than longer words in the perspective conditions. This hypothesis was supported through results indicating that participants in each condition more easily recalled short words compared to long words. The hypothesis that familiar words, when compared to those of the same length in the unfamiliar condition, would be recalled in the correct order with greater accuracy than unfamiliar words was also supported. Results indicated that participants in the familiar condition recalled both long and short words at significantly greater rates than did those in the unfamiliar condition. These results are consistent with those of previous research (Baddeley et al., 1975; Cowan, Wood, Nugent, & Treisman, 1997; Hulme et al., 2006; Roche et al., 2011).
The researchers of this study also found that participants tended to remember either the first couple of words or the last couple of words from the recordings at greater rates than the middle words. This result matches the primacy-recency effect phenomenon, also researched by Murdock (1962) who asked participants to write down as many words as they could remember in no particular order after hearing a set of words read aloud. Murdock (1962) found that participants were most likely to recall the last and first words from the set of words read aloud rather than middle words. Since the participants in our study were asked to recall in order, it is possible that if not asked to recall in order they would have recalled even greater numbers of words and more of the last words. While the hypotheses of the present study were dependent on participants recalling words in the correct order, some participants recalled a greater number of words correctly overall. However, statistical analyses indicated that the same effects of familiarity and word length were present in the overall recall of correct words, again indicating that participants had greater rates of recall for familiar words compared to unfamiliar words and short words compared to longer words. Applying these results to the real world, a study by Tyson & Kramer (1981) found that counselors responded to content rather than the positioning of messages when listening to clients. Again, this further illustrates that the primacy-recency effect may not be the sole indicator of participants’ ability to recall information but that importance and familiarity of content also play an essential role.
Looking back at this study, there were several limitations. As with many studies using a college participant pool, a major limitation of this study was the lack of participants for each condition. Age was also a limitation as participants were between the ages of 18 and 23. Since age has an effect on the capacity of memory, it would be important for further research to look outside the college participant pool to account for any age differences that may be present. Gender was evidently another limitation in that the majority of the participants of the present study were women. As this study pertained to the Connecticut College community and used words from locations around the campus as familiar words, an additional limitation was that the study could have only been done by students from this community and the familiarity of these words may have had a different level of effect on the participants than other familiar words could have had. Lastly, since the researchers of this study decided to use the same words in each recording following the methods of Baddeley et al. (1975), Hulme et al. (2006) and Roche et al. (2011), the present study did not account for the effect of repetition on memory recall. Tussing and Greene (2001) examined the effect of familiarity and repetition on memory recall and found that the accuracy of recall of words was higher when the list contained repeated words compared to a list that contained distinct words. Since the researchers averaged the results of the five trials, hopefully
the possibility of repetition recall was accounted for. Even with the measures taken in the current study, this is a subject that should be further explored in the future.
Along the lines of further research on repetition, another potential topic could be to examine the effects of rehearsal on memory recall. This suggestion was also brought forward by McNeil and Johnston (2004) in a study in which they examined word length effect and the effect of having visual stimuli with words for poor and normal readers. As the current researchers concur, future research should take into account other possibilities of factors that could influence the word length effect such as repetition and rehearsal. Another future direction for research on word length effect and familiarity would be to account for gender differences. A study was conducted by Kaushanskaya, Marian, and Yoo (2011) in which they compared the ability of men and women in recalling phonologically-familiar novel words and phonologically-unfamiliar novel words. They used words that were made-up but differed in that the phonologically-familiar novel words were made up of phonemes found in the English language and phonologically-unfamiliar novel words were constructed using phonemes found in non-English languages. Kaushanskaya et al. (2011) found that women were significantly better at recalling phonologically-familiar novel words than men, however there was no difference on performance in recalling phonologically-unfamiliar novel words. As the present study did not have enough participants from each gender to account for any possible gender differences in familiarity, and the study by Kaushanskaya et al. (2011) only accounted for familiarity, not the word length effect, future research could expand upon such gender differences combining research on familiarity and word length effect.
As previously mentioned in research by Stuart & Hulme (2000), the effects of familiarity, due to long-term associations, play an important role in short-term word memory tasks. Tehan and Tolan (2007) also found that the word length effect can pertain to long-term memory in addition to short-term memory. As noted by Tehan and Tolan (2007) as well as Hulme and Maughan (1991), the word length effect has often been ignored in long-term recall and is a direction future studies could take. Again, it would be interesting to combine both familiarity and the word length effect, as did the present study, to further examine the role these factors have in long-term memory. Finally, the present study did not compare the effects of participants’ majors, which could influence recall abilities through familiarity and word length effect. For example, it could be possible that students who are required to memorize information for their majors might be better at memorizing words in this study.
Most importantly, the present found that familiarity and word length effect both have an influence on the recall of words in short-term word memory tasks. Such findings are important to help us understand why people remember some information more readily than other information. The researchers hope to take into account limitations and potential further research in a future study expanding on the present findings on the effect of word length effect and familiarity.
Aly, M., Yonelinas, A. P., Kishiyama, M. M., & Knight, R. T. (2011). Damage to the lateral prefrontal cortex impairs familiarity but not recollection. Behavioural Brain Research, 225, 297-304. doi:10.1016/j.bbr.2011.07.043
Anderson, J. R. (2009). Cognitive psychology and its implications. (7th ed.). New York: Worth.
Baddeley, A. D., Thomson, N., & Buchanan, M. (1975). Word length and the structure of short-term memory. Journal of Verbal Learning& Verbal Behavior, 14, 575-589. doi:10.1016/S0022-5371(75)80045-4
Cowan, N., Wood, N. L., Nugent, L. D., & Treisman, M. (1997). There are two word-length effects in verbal short-term memory: Opposed effects of duration and complexity. Psychological Science, 8, 290-295. doi:10.1111/j.1467-9280.1997.tb00440.x
Hulme, C., & Maughan, S. (1991). Memory for familiar and unfamiliar words: Evidence for a long-term memory contribution to short-term memory span. Journal of Memory and Language, 30, 685-701. doi:10.1016/0749-596X(91)90032-F
Hulme, C., Neath, I., Stuart, G., Shostak, L., Surprenant, A. M., & Brown, G. D. A. (2006). The distinctiveness of the word-length effect. Journal of Experimental Psychology: Learning, Memory, and Cognition, 32, 586-594. doi:10.1037/0278-7318.104.22.1686
Jalbert, A., Neath, I., Bireta, T., & Suprenant, A. (2011). When does length cause the word length effect? Journal of Experimental Psychology: Learning, Memory, and Cognition, 37, 338-353. doi:10.1037/a0021804
Kaushanskaya, M., Marian, V., & Yoo, J. (2011). Gender differences in adult world learning. Acta Psychologica, 137, 24-35. doi:10.1016/j.actpsy.2011.02.002
Kolic-Vehovec, S., & Arar, L. (2005). Effects of surname familiarity on recall. Studia Psychologica, 47, 35-50.
McNeil, A. M., & Johnston, R. S. (2004). Word length, phonemic, and visual similarity effects in poor and normal readers. Memory & Cognition, 32, 687-695. doi:10.3758/BF03195859
Miller, G. A. (1956). The magical number seven, plus or minus two: some limits on our capacity for processing information. Psychological Review,63, 81-97. doi:10.1037/h0043158
Murdock, B. B. (1962). The serial position effect of free recall. Journal of Experimental Psychology, 64,482-488. doi:10.1037/h0045106
Richards, L. G., & Heller, F. P. (1976). Recognition thresholds as a function of word length. The American Journal of Psychology, 89, 455-466. doi:10.2307/1421618
Roche, J., Tolan, G.,& Tehan, G. (2011). Concreteness effects in short-term memory: A test of the item-order hypothesis. Canadian Journal of Experimental Psychology-Revue Canadienne de Psychologie Experimentale, 65, 245-253. doi:10.1037/a0024693
Saint-Aubin, J., & Poirier, M. (2005). Word frequency effects in immediate serial recall: Item familiarity and item co-occurrence have the same effect. Memory, 13, 325-332. doi:10.1080/09658210344000369
Stuart, G., & Hulme, C. (2000). The effects of word co-occurrence on short-term memory: Associative links in long-term memory affect short-term memory performance. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 796-802. doi:10.1037//0278-7322.214.171.1246
Tehan, G., & Tolan, G. A. (2007). Word length effects in long-term memory. Journal of Memory and Language, 56, 35-48. doi:10.1016/j.jml.2006.08.015
Turner, J. E., Henry, L. A., & Smith, P. T. (2000). The development of the use of long-term knowledge to assist short-term recall. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 53A, 457-478. doi:10.1080/027249800390565
Tussing, A. A., & Greene, R. L. (2001). Effects of familiarity level and repetition on recognition accuracy. The American Journal of Psychology, 114, 31-41. doi:10.2307/1423379
Tyson, G. A., & Kramer, D. (1981). The Primacy-Recency effect in counselor response. Journal of Clinical Psychology, 37, 88-90. doi:10.1002/1097-4679(198101)37:1<88::AID-JCLP2270370115>3.0.CO;2-I
van Kesteren, M. T. R., Ruiter, D. J., Fernandez, G., & Henson, R. N. (2012). How schema and novelty augment memory formation. Trends in Neurosciences, 35, 211-219. doi:10.1016/j.tins.2012.02.001
Wilson, M., & Emmorey, K. (1998). A 'word length effect' for sign language: Further evidence for the role of language in structuring working memory. Memory & Cognition, 26, 584-590. doi:10.3758/BF03201164