Wingfield, et al (1988) found that younger subjects performed better than older ones in tasks that demanded simultaneous storage of recently presented material, and processing of additional information. In the area of explicit and implicit memory, it was found that younger subjects are superior in explicit memory tasking that require intention to remember. As in the study demonstrated by Kausler and Hakami (1983), it was found that memory for activities that one was involved in, is a rehearsal-independent form of episodic memory.
Despite being rehearsal-independent, they found that younger adults performed better. In support to these findings, Light and Singh (1987) reviewed the study of Kausler and Hakami (1983) and their results supported their findings that older adults do best when asked to recognize words they learned earlier and somewhat worse when asked to recall words using a meaningfully-related cue. An explanation for this is that, although encoding of activities is an automatic process, retrieval from long-term episodic store may be effortful for the elderly (Kausler and Lichty, 1988).
According to Belsky’s (1990) review of findings, deficits can occur at the encoding, storage, or retrieval stages of information processing – “Both retrieval and encoding deficits are implicated in the lower performance of older adults… it is not likely that… deficits can be isolated to one stage or component, but rather that deficits are widespread in the entire cognitive system. ” (p. 443). However, the older person does not necessarily show deficits in all aspects of memory.
While age differences are negligible for implicit memory performance, the latter does not involve conscious recollection of event occurrence (Light and Singh, 1987) (see also, Kausler 1991). Explanations for the decline in memory functioning as one ages, shall vary from some basic physiological process like the death of brain cells, to the declination in physiological arousal and mental activation (Glass and Holyoak, 1986). An alternative view offered by Houx, et al. 1991) is that many age effects reported in the literature can be largely explained by suboptimal brain functioning which is not related to age. A summary of the various models of memory outlined by Klatzky (1990) in her article on theories of information processing and ageing, pointed out that the models of memory assumed a structure for representing information about the external world, ideas, and internal states. And memory systems can be conceptualized in several ways, including short-term versus long-term memory, semantic versus episodic memory, and automatic versus effortful processing.
In the automatic processing, it requires negligible amount of attention (for example, estimating the number of times an item was encountered, recalling location of objects and duration of events). In contrast, effortful processing requires conscious attentional effort towards encoding or retrieval of information (as in remembering novel names). Automatic Processing Hasher and Zacks (1979) proposed a framework for conceptualizing a wide range of memory phenomena. Based on the results of four experiments conducted by them, several characteristics that distinguished automatic and effortful processing were outlined.
The basic assumption is that encoding processes differ in the amount of attention required, such as automatic processing uses little or no capacity, occurs without intention, and is an innate ability. In contrast, effortful processing uses attention, occurs with conscious effort, and is dependent on strategy and practice. These two processes can be best thought of as belonging to a continuum of attentional requirement. Hasher and Zacks (1979, 1984) also proposed the criteria for evaluating their framework and the predicted results.
First, individuals should perform equally well in automatic tasks (frequency-of-occurrence, spatial, and temporal information), whether the information was incidentally or intentionally learnt. Second, automatic processing is not susceptible to instruction or practice, while effortful processing is dependent on the type of instruction given, and greatly enhanced with practice. Third, automatic processes are not affected by task interference, but effortful processes will interfere with each other when they exceed the amount of available capacity.
Fourth, depression or high arousal affects the performance on effortful tasks, but has no bearing on automatic processing. Finally, automatic processing shows negligible change across different ages, while effortful processes are performed less well by young children and the elderly. Strayer and Kramer (1990) outlined an additional “zero slope criteria”. Performance of automatic encoding tasks as measured against manipulated variations in the levels of any independent variable should yield no change. In contrast, effect of changes in the independent variable can be observed in effortful memory performance.
And many researchers subsequently tested Hasher and Zack’s (1979) framework, yielding results that either verified or refuted one or more of the 5 evaluative criteria. Of interest, to their study, they claimed that intentional learning will not affect automatic processing, and that performance is independent of age variations – suggesting that when an individual encodes information without conscious awareness of doing so, the information is incidentally learnt. However, their views are being raised by Ellis and Rickard (1989) regarding the durability of automatically encoded data.
Thus, they assessed memory over 24-hour retention interval, and found that automatically processed information is encoded into long-term memory. They also verified that incidental and intentional instructions did not cause any significant differences in response. Another study yielded results that are both supporting and refuting towards Hasher and Zack’s (1979) claims. Greene (1990) showed that subject’s memory was above chance level even at the absence of anticipation of a test, demonstrating that there was sensitivity to frequency information without intention.
However, it was also found that the strategy employed influenced performance accuracy, as opposed to Hasher and Zack’s (1979) prediction. Naveh-Benjamin and Jonides (1986) demonstrated that encoding of frequency information is adversely affected by increasing the load of a competing task demand, and that a specific strategy manipulation affected frequency encoding. Subjects who generated semantic associates of stimulus words gave more accurate frequency estimates than did subjects who only generated acoustic associates.
Furthermore, Jonides and Naveh-Benjamin (1987) found that deeper processing of stimulus material resulted in significantly superior frequency judgment than did shallower levels of processing. Maki and Ostby (1987) concluded from their study that frequency information is not encoded automatically because it is affected by the level of attention in the initial encoding stage, and with deeper levels of processing, it improved performance on frequency task – suggesting that frequency encoding may be susceptible to attentional allocation.
Based on the above reviews, it is observed that, depending on the depth of information processing which is manifested through incidental or intentional learning of stimuli, performance on frequency estimation varies. Kausler (1991) discussed frequency-of-occurrence memory in relation to aging. There is generally no major decline in the proficiency of frequency-of-occurrence memory from early to late adulthood. However, distinction should be made between absolute and relative frequency estimations.
Although modest age-effect is found when the test is of absolute frequency, relative frequency estimation tests yielded no significant age variations. Hence, is frequency-of-occurrence estimation vulnerable to the effects of intention and age? There are two ways towards the resolution of this issue. One, consideration is the involvement of the working memory, and the other concerns the cover tasks typically employed by researchers in the manipulation of the incidental condition. Working Memory Capacity (Effortful Processing)
In effortful processing (like remembering the contents of a free recall list), performance is dependent on working memory capacity. Since this declines with age, older persons will show lower performance. However, automatic processing (as in encoding of frequency information) occurs without conscious effort, and is thus not affected by any decline in working memory capacity. And according to Kausler (1991), there is no age effect as long as working memory is not involved in the process. Thus, this conceptualization of automaticity provides a preliminary indication of how discrepancies between findings can be resolved.
Contrary to Hasher and Zack’s (1979) claims, several studies (e. g. Jonides and Naveh-Benjamin, 1987; Maki and Ostby, 1987) found automatic processing to be vulnerable to the effects of intentional learning. Proposing that it is possible that the tasks employed in the various studies tapped the working memory capacity by varying degrees, therefore giving rise to differences in results. To reconcile the researcher’s conflicting claims on the automaticity of frequency processing, Sanders, et al. (1987) proposed the non optimal model for frequency memory.
According to this model, frequency encoding uses very little effort and working memory space, for performance to be above chance level. In task conditions that tap only automatic processing, age differences between young and old should be negligible. However, under intentional learning conditions, a deeper level of processing becomes effortful, which renders it vulnerable to the limits of one’s working memory capacity. And as capacity of working memory declines with advancing age, deficits in performance may result. Cover Task
In order to create an incidental learning condition, subjects need to be engaged in a cover task that will draw their attention to the stimuli, and yet do not result in purposeful encoding of item attributes. In relations to this experiment of Sanders, et al. (1987), Johnson, et al. (1989) found that judgment of relative frequency was affected by the nature of the cover task subjects engaged in while viewing the stimuli and improved frequency judgment was found when subject’s attention was directed to the identity of the stimuli (semantic processing).
However, Kausler (1991) pointed out that many of the incidental learning conditions employed in frequency-judgment research is not, incidental as subjects are often told to expect a memory test, nature of which is unspecified. Thus, undoubtedly, they will engage in some form of rehearsal, and “age differential in the extent to which items are rehearsed under such conditions (are) sufficiently so to mask any age difference in the proficiency of frequency-of-occurrence memory. ” (p. 433). Therefore, to avoid this problem, subjects need to be totally ignorant of any upcoming memory test.
And age difference failed to reach statistical significance in Sanders et al. ‘s (1987) task, whereby subsequent administration of a memory test was disguised – suggesting that in the present study, subject’s attention should be draw to the non-content attributes of the stimuli (to avoid semantic processing), and they will not be informed of subsequent memory test in the incidental condition. Future Directions In all the studies just reviewed, a future research is needed to examine the concept of automaticity as characterized by Hasher and Zacks (1979).
And automatic processing will be continue to examine across two of their evaluative criteria outlined – that automaticity is unaffected by age differences (younger versus older adults), and mode of learning (incidental versus intentional learning conditions). And so does their interaction effects between the two independent variables that intention is expected to enhance the performance of both age groups. The present experimental design proposed, should be using a cross sectional studies.
The obvious limitation of such studies is that developmental trends cannot be inferred, unlike longitudinal studies which will provide results indicating changes in relation to growth and maturation. Furthermore, whether age effects found are due to cohort difference is questionable. However, Hartley (1980) noted several merits of cross-sectional research over longitudinal studies. First, in the area of research on learning and memory, there are rapid changes in theoretical and empirical orientations – suggesting that methods used in longitudinal studies may become obsolete across years.
Second, when there are differences found in different cohorts in cross-sectional designs, there is the opportunity to find theoretical and empirical explanations to account for the findings. Differences due to cohort can be evaluated by considering the variables inherent in the different groups of subjects. And Hartley considers it more important to identify the mechanisms behind the processes than to classify the causes of differential performance.
The implication for the proposed studies is to identify age-related factors that cannot be partialed out, and to consider their impact on memory performance. According to Krauss (1980) discussion on between-and within-group comparisons in aging research, the 3 common confounding variables are (a) unequal ration of each gender of subjects in the different groups; (b) health status; (c) unbalanced educational levels between the younger and the older subjects. There is often unequal are ranges in the two groups of subjects used.
Thus, all these factors should be taken into consideration when performing the design of the present study. In general, the many studies which refuted the conceptualization of Hasher and Zacks (1979) reflected the functionalist perspective that there are individual variations between people. Bruce (1990) in his discussion on functionalist explanations of memory commented that Hasher and Zack’s claims are at odds with the fundamental postulate of the functional approach, which maintains that there is diversity within a population.
As opposed to supporting the null results predicting no age difference (as in many automatic processing research), thus, the present study should hypothesize the significant effects. Frequency of occurrence is selected as the information to be encoded and recalled, despite some dispute over whether it is automatically processed. Some research has provided support for frequency-estimation as an automatic process (cf. Sanders et. al, 1987 and Kausler, 1991), while others have not (cf. Naveh-Benjamin and Jonides, 1986/7).
Essentially, the question that needs to be address is the extent to which frequency information is automatically processed based on the proposal of Sanders et al. (1987) in the view of automatic processing that need not produce optimal output, and Naveh-Benjamin (1987) that first, performance on frequency encoding may not be as dependent on intentional learning as free recall of names and second, innate mechanisms which assist people in operating functionally in their environment can be considered as automatic.
Hence, the study should also provide an opportunity to rectify previous methodological flaws in other research. Proposing that first, care should be taken in the design of the study and selection of subjects, such that extraneous cohort factors will be controlled for. Second, manipulation of the identical learning condition should be strictly so by the use of an appropriate cover task, in which subjects do not anticipate any memory test (as in the case of the study conducted by Kausler, 1991), and thus will not engage in any purposeful encoding of stimuli in the incidental learning condition.
As Kausler noted that knowledge about frequency of occurrence in the real world more often pertains to categorical knowledge rather than discrete events. For instance, people may be aware that there are more comedies than documentaries on television, without keeping count of each individual program. Hence, in future, test lists consisting of different categories could be used. For example, a list could contain pictures of different fruits, varieties of vehicles etc.
Subjects could then be tested on which category had more items in the list, instead of frequency judgment on discrete stimuli. And one criticism of Hasher and Zack’s (1979) experiment is that although automatic processing was investigated, effortful processing was not tested against their evaluative criteria. Therefore, the present study should also include an investigation on effortful processing, using the same set of stimulus material employed for the automatic task – suggesting that the overall study research proposal consist of two experiments with two separate groups of subjects.
One, examining effortful processing based on the hypotheses that (a) younger subjects will recall more names than older subjects, and (b) performance will be enhanced for both age groups when stimuli encoded intentionally than identically. And lastly, automatic processing, in which subjects give frequency-of-occurrence information (dependent variable) based on the hypotheses that (a) younger subjects will give more accurate frequency estimations than the elderly, and (b), performance will be enhanced for both age groups when frequency information is encoded intentionally than incidentally.
In order to conduct this research proposal, cross-sectional method of age comparison seemed to be a preferable design method having reviewed (cf. Hartley, 1980); that is a group of young adults (often college students in their early twenties) is compared with a group of older adults, usually volunteers in their sixties and seventies. They do this by matching the groups by educational level, vocabulary, and by other indicators of intelligence and socioeconomic status.
Many crucial experimental results take the form of interactions between age and some experimental variable; that is, one condition of the experiment is associated with large are-related differences, whereas another condition is associated with much smaller differences even though the same subjects are used. And three steps of techniques in this methodology were suggested by Hartley (1980). First, researchers could generate an exhaustive list of everyday cognitive behavior.
Second, the varying lengths of time different group pf people (across age, gender, occupation, etc. ) spend on each type of behavior should be measured. Finally, these behaviors could be classified in terms of the extent of cognitive abilities required. Thus in future, these daily activities could be employed in the investigating of automatic and effortful processing. The use of daily tasks would not only have higher face validity, especially for the elderly, but greater ecological validity as well.