i

The Journalof GeneticPsychology, 2005, 166(3), 313-327

Feature Binding in Children and Young Adults THOMAS C. LORSBACH Department of Special Education and CommunicationDisorders University of Nebraska at Omaha JASON F. REIMER Department of Psychology CaliforniaState University, San Bernardino

ABSTRACT. The authors measured memory for individual features (objects only or locations only) and the combination of those features (objects and locations) in 9-, 12-, and 21-year-old students with a yes or no recognition task. Analysis of recognition memory performance (d' scores) revealed that although age differences existed in memory for individual features, age differences were greater in the tasks that required memory for combined features (objects and locations). Hierarchical multiple regression analyses indicated that age remained a significant predictor of memory performance in the combination condition even after the authors statistically removed memory performance in object and location conditions and the interaction effects of object and location. These results provide evidence for developmental differences in the binding of features in memory. Key words: binding, children, feature memory THEORISTS HAVE CONCEIVED OF MEMORY REPRESENTATIONS'as collections of features or attributes (Bower, 1967; McClelland & Rumelhart, 1985; Underwood, 1969). Depending on the nature of the episodic event and which stimulus features are encoded, a given memory representation may consist of a variety of features, which range from surface elements to deeper, semantic features. Consistent with the position that memory representations are composed of a collection of attributes, researchers have observed that memory retrieval is often fragmentary, with certain features being remembered and others unavailable or inaccessible. For example, a person may be able to remember certain elements of the representation, such as the location in which information was displayed on a page, but be unable to retrieve the information itself (ZechAddress correspondence to Thomas C. Lorsbach, Department of Special Education and Communication Disorders, University of Nebraska at Omaha, 6001 Dodge Street, KH 421C, Omaha, NE 68182-0054; [email protected](e-mail). 313

314

The Journal of Genetic Psychology

meister & McKillip, 1972). Conversely, a person may be able to remember the information but be unable to remember when or where the information was acquired. Retrieval of the information without memory for supporting details is typical of failures in source memory (Johnson, Hashtroudi, & Lindsay, 1996). The features that comprise a given memory episode aie not stored in a random manner, rather they must be bound together so that they form a unique representation of the event. Researchers refer to the process by which features of an episode are associated or connected as binding or cohesion. Chalfonte, Verfaellie, Johnson, and Reiss (1996, p. 610) observed that binding "provides the memorial experience that features belong together,"Kroll, Knight, Metcalfe, Wolf, and Tulving (1996, p. 178) considered that cohesion occurred early during consolidation, a process "whose function is to bind or glue aspects of incoming information into separately retrievable engrams." Metcalfe, Mencl, and Cottrell (1994) observed that connecting features of a memory episode into a bound representation would distinguish explicit from implicit memory. Metcalfe et al. considered memory-binding processes to be a "crucial characteristic" of the explicit memory system because they "coalesce the separate parts of an event into a cohesive and memorable whole" (p. 392). Without adequate feature binding, explicit memory may be compromised and may ultimately lead to source memory failures in which fragmentary information of an episode is remembered without a cohesive memory of where and when the information was acquired (Schacter, Norman, & Koustaal, 1998). Researchers have not yet addressed whether feature binding shows developmental improvement. Therefore, in this study, our purpose was to 'examine whether developmental differences exist in the ability to bind features together in a working memory task. For the present study, we adapted the K. J. Mitchell, Johnson, Raye, Mather, and D'Esposito (2000) procedure to examine feature memory and binding processes in children. -We presented participants with alternating blocks of trials that included study and test items for each of three memory conditions: (a) item only, (b) location only, or (c) item and location. We began each trial with a series of three 3x 3 study arrays in which each array contained a different line drawing in a different random location. Before receiving each block, we instructed each participant to remember only the items, to remember only the locations of items, or to remember both the items and their locations. Following a brief unfilled retention interval, we gave participants a yes or no recognition test to assess their memory for individual features (item-only information or location-only information) or memory for combined features (item and location information). In at least two studies, researchers used a small-scale grid or matrix to compare the development of item and location memor6 in children (Kail & .Siegel, 1977; Siemens, Guttentag, & McIntyre, 1989). In each case, investigators were interested in memory for occupied locations (i.e., memory for those occupied locations on the matrix, regardless of the items that had occupied those locations).

Lorsbach & Reimer

315

Kail and Siegel presented letters in a 4 x 4 matrix to third-grade students, sixthgrade students, and college students with instructions to remember only the letters, only the occupied locations, or the letters and their locations in the matrix. Although they found memory performance on both the letter- and the locationmemory tasks to increase with age, the scoring method did not reveal anything about binding processes because memory for letters and their locations were scored independently (i.e., memory for letters was scored without regard to the original location in the matrix and vice versa). Siemens et al. used a modified version of the Kail and Siegel task to compare memory for items with memory for occupied locations in children (4-year-olds, 8-year-olds) and adults. Age differences in memory performance were significantly greater on the location-only memory task than on the item-only memory task. Unfortunately, Kail and Siegel (1977) and Siemens et al. (1989) did not provide information about age differences in memory for combined features (objects and their locations). However, because the combination task required participants to bind objects with their locations, developmental differences in the combination condition should have been significantly greater than those that may have been observed in the feature conditions (i.e., object and location). We based such an expectation on what we knew about the nature of feature binding as well as our knowledge of memory development in children. For example, researchers have found that limitations on attentional resources adversely affect the efficiency of binding processes (e.g., Kroll et al., 1996; Reinitz, Morrissey, & Demb, 1994). Although developmental models do not typically portray children as possessing fewer resources than adults, children have been considered to be slower and less efficient than young adults in the execution of mental processes (Case, 1985; Case, Kurland, & Goldberg, 1982). Because of general processing inefficiency, children may be able to process individual features (objects or their locations) but may have fewer resources left to bind these features together during the combination task. A second reason to anticipate developmental differences in binding processes comes from research in developmental neuropsychology. The hippocampus and the prefrontal cortex seem to be the neurological locus of binding processes (e.g., Cohen & Eichenbaum, 1993; Eichenbaum & Bunsey, 1995) and the frontal lobes show neurological development until the adolescent years (Huttenlocher, 1990; Johnson, 1999; Yakovlev & LeCours, 1967). One might expect to find developmental improvements in a working memory task that requires binding processes. A final reason to expect developmental differences in the combination condition is in relation to the research on adult aging in which investigators have found binding processes to be impaired in older adults (Chalfonte & Johnson, 1996; K. J. Mitchell, Johnson, Raye, & D'Esposito, 2000; K. J. Mitchell, Johnson, Raye, Mather, & D'Esposito, 2000). The results of adult aging studies often are useful in generating hypotheses about cognitive development because a num-

316

The Journalof Genetic Psychology

ber of cognitive processes that decline with adult. aging have been found to improve as children develop. These processes include cognitive inhibition (e.g., Dempster,'1992; Hasher & Zacks, 1988), speed of processing (e.g., Kail, 1991; Cerella, 1985), and implicit versus explicit memory (e.g., D. B. Mitchell, 1991; Parkin, 1991). Therefore, because binding processes-seem to decline as the result of adult aging, these same processes may improve throughout the course of child development. Method Participantsand Design

The participants included 26 third-grade students (M age = 9.24 years, SD = 0.35), 27 sixth-grade students (M age- 12.07 years, SD = 0.46), and 27 college undergraduate students (M age = 21.11 years, SD 3.4). We used a 3 x 3 (Grade [3rd x 6th x college] x Memory Condition [object x location x combination]) mixed design, in which grade represented the between-subjects variable and the memory condition represented the within-subjects variable We recruited college students from undergraduate psychology classes at the University of Nebraska at Omaha through the use of sign-up sheets. We recruited children from two public elementary schools in the Omaha, NE, metropolitan area by sending a letter of invitation and a parent consent form to all parents of children in the third and sixth grades. We excluded from participation children who did not speak English as their native language or who were currently receiving special education services. Participation in the experiment was contingent on the'completion of signed consent forms and child assent forms, which had been approved by the Institutional Review Board for the Protection of Human Subjects at the University of Nebraska Medical Center. Apparatus and Procedure

We presented stimuli and collected the responses'by using a Power Macintosh 6100/60 AV microcomputer that was controlled by SuperLab Pro 1.75 software (Cedrus Corporation, 1989-1999). We interfaced an RB410/RB 610 response box (Cedrus Corp.) with the microcomputer to record both the accuracy and latency of each response. We presented stimuli on a 14-in color monitor. The critical stimuli consisted of a set of eight black-and-white line drawings of common objects (pumpkin, fish, balloon, kite, snowman, lion, frog, heart) and a black 3 x 3 grid that was approximately 15.5 cm x 15.5 cm. We used only eight pictures to equate the number of objects tested with the number of locations tested (i.e., we used only 8 locations because we never used the center cell of the 3 x.3 grid as a study or a test location). The pictured objects did not possess any obvious categorical relationships. In addition, children's ratings (Cycowicz,

Lorsbach & Reimer

317

Friedman, & Rothstein, 1997) of the pictured objects in terms of their familiarity (M = 2.84, SD = 0.43), visual complexity (M = 2.78, SD = 0.94), and percentage of name agreement (M = 98%, SD = 3.51%) were similar to those of young adults (M = 2.85, SD = 0.55; M = 2.75, SD = 1.1; and M = 98.9%, SD = 2.47%, respectively) in the norms of Snodgrass and Vanderwart (1980). We initially constructed three 32-trial test lists for each of the three memory conditions: (a) memory for objects, (b) memory for locations of objects, and (c) memory for objects and their respective locations (combination) on the grid. We selected the eight pictures as study objects through repeated and random selection without replacement, in which we used the eight pictures as study items on each test list. Each picture was used the same number of times. The test probes in each list consisted of 16 targets and 16 lures. We presented the targets and lures randomly with the restriction that no more than three targets and three lures were tested consecutively. We tested the eight study locations of targets twice, and we tested the three study positions (1st, 2nd, 3rd) equally often with each test list. We varied the order of testing of the three memory conditions systematically across participants so that the three conditions were tested in each third of the experimental session equally as often with each age group. In addition, across the three test lists used within a given experimental session, we tested each of the eight item locations twice at each of the three study positions. Figure 1 graphically depicts the sequence and timing of stimulus events used in each trial. With the exception of the study and test cues, we presented all stimuli as black objects on a white background. We signaled the beginning of each trial by displaying the word study, which we presented in green letters in the center of the screen for 500 ms. A blank screen then was displayed for I s and was followed by the 3 x 3 grid that remained visible for 3 s. During this 3-s interval, we showed three different pictures successively for 1 s each in three different locations in the grid. An unfilled 8-s retention interval followed the presentation of the third object on the grid, during which time we showed a question mark (?) in the middle of the screen. At the end of the 8-s retention interval, we displayed the word test in red letters for 500 ms, which signaled that a test item was about to be shown. A blank screen then was displayed for 1 s and was followed by the test item for that trial. The test item remained visible until the participant responded by pressing one of two buttons that were labeled either yes or no. The participant's response was followed by a 2-s intertrial interval. Depending on the condition being tested, the recognition probe consisted of a single item that tested memory for objects, locations of objects, or objects and their respective locations (combination). For object memory trials, the test item consisted of a single object that was presented in the center of the screen. We instructed participants to press the yes button if they thought the object was one of the three studied items that had been displayed on that trial and to press the no button if otherwise. On location memory trials, the test item consisted of the grid containing a single darkened cell (excluding the center cell). We instructed the

318

The Journalof Genetic Psychology

.2

0 0

a)

o•+L) ,)

a) C'-

0

a)

o

. a)

C

v-

a)

Is .0 0.

W

•

•0

0 0.

a)

.0 .0 0 .0

0

'0a)

U2

0

0.

a)

*0

U,

a)

.0 a)

*0 0

0

0 kn

CO

a) a). C) 0)

a)

.0

.0

C.-

0

0 0 'U CU,

-p.a) 'U 00

Cj2

0)

0 0

0 0

C0 0

0 0

00

a)

.0

a)

Lorsbach & Reimer

319

participants to press the yes button if they thought the darkened cell was in a location in which one of the three studied items had been displayed on that trial and to press the no button if otherwise. For combination trials, the target item consisted of an object in its studied location on the grid for that trial. The lures for this condition involved the re-pairing of objects and their respective locations. We instructed participants to press the press the yes button if the test item displayed an object in its studied location on the grid and to press the no button if otherwise. In all test conditions, participants were instructed to make their decision as quickly as possible without sacrificing accuracy. We used an 8.5 x 11 in (21.5 x 28 cm) sheet of paper, which contained an example of the 3 x 3 grid and the eight pictures as a visual aid during the presentation of oral instructions. We told participants that on each trial, three different pictures would be displayed in a random manner, one at a time, in three different squares within the grid. In addition, just prior to each memory test, we presented test-specific instructions through the use of another visual aid. We showed participants a 6-page, 8.5 x 11 in (21.5 x 28 cm) booklet that provided an example of a stimulus sequence that might occur during the trials of a given memory test. Following the specific instructions for a given memory task, the participants completed two practice trials to familiarize them with the task and its requirements. We presented participants with the test trials only after they demonstrated their understanding of the procedure as assessed during practice trials. Results We initially computed hit rates and false-alarm rates for all participants in each of the three memory conditions. We determined hit rates by calculating the proportion of times the participant responded by pressing the yes button when the probe item was displayed in the preceding study grid (i.e., a correct response). We determined false-alarm rates by calculating the proportion of times the participant pressed the yes button when the probe item was not displayed on the previous study grid (i.e., an incorrect response). We subsequently used the hit and false-alarm rates to compute d', a measure of memory discriminability, and C, a measure of response bias (see Stanislaw & Todorov, 1999, for an overview of signal detection theory and methods for calculating d' and C). The measure of d' assesses the ability to discriminate between old and new items, with larger d' values reflecting better memory discrimination. Values of C that are below 0.0 are indicative of a liberal bias (i.e., greater willingness to guess yes), whereas values above 0.0 suggest a more conservative bias (i.e., less willing to guess yes). Before calculating d' scores, we transformed the hit and false-alarm rates for each participant using a log-linear rule (Hautus, 1995). Such a correction of hit and false-alarm rates made extreme proportions impossible to obtain and has been found to produce less biased estimates of d' (Hautus). Likewise, we corrected hit and false-alarm rates by adding .5 to each frequency and dividing by N + 1, where

320

The Journalof GeneticPsychology

N equals the number of old or new trials (Snodgrass & Corwin, 1988, p. 35). Thus, we calculated corrected hit rates and false-alarm rates using the following formulas: (a) hits = (number of hits + ;5) (number of old items + 1)and (b) false alarms = (number of false alarms + :5)

(number of new items + 1). Snodgrass

+

and Corwin recommended that this correction procedure be used routinely, even in the absence of extreme scores and even when signal detection measures are not calculated. Table 1 shows the mean proportions of corrected hits and false-alarms for each grade level and memory condition. Using corrected hit and false-alarm .rates, we calculated the d' scores for each participant and submitted to a 3 x 3 (Grade [3rd x 6th x college] xMemory Condition [object x location × combination]) mixed-design analysis of variance. The analysis of d' scores revealed that the effects of grade, F(2, 77) = 32.740, MSE= .822, p .14). We performed an analysis of response bias on C values. Values of C that were less than 0.0 reflected a liberal response bias (i.e., greater willingness to respond yes), whereas C values greater than 0.0 indicated a more conservative bias (less willing to respond yes). Only the effect of memory condition was significant, F(2, 154) = 5.970, MSE = .075, p = .0032. The C values in the combination memory condition (M = -. 04) differed from both the object memory (M = .10) and the location memory condition (M = .07), but the latter two did not differ from each other. Thus, all participants were relatively more conservative when making recognition decisions with both the object and the location memory conditions and were somewhat more liberal in the combination memory condition. In the present investigation, our purpose was to determine whether developmental differences exist in memory for bound featural information. Although we found developmental differences in the combination condition, interpretation of these differences is difficult because we also found age differences in feature memory. In the future, researchers need to examine whether developmental differences can be found in the combination condition independent of memory performance in the featural conditions (i.e., object and location). To be specific, we were interested in determining whether developmental differences in the combination condition remained after variance associated with featural memory had been removed. To address this issue, we conducted a hierarchical multiple regression analysis. In the first step of the regression analysis, we entered memory performance in both the object and location conditions. In the second step, we entered an interaction term representing the product of memory performance in the object and location conditions. In the third step, we entered the main effect of age in the form of two contrast coded variables, one examined differences in combination memory performance between third- and sixth-grade students, and the other focused on differences in combination memory performance between both third- and sixth-grade students and college students. On the first step, combination memory performance could be significantly predicted from object and location memory performance, R = .765, R2 = .585, F(2, 77) = 54.26, p < .01. Object memory, unstandardized B = .615, [3 = .448, t(79) = 5.10, p < .01, and location memory, unstandardized B = .477, f = .421, t(79) = 4.79, p < .01, were each positively associated with performance on the combination condition. We entered the interaction of objects and location memory performance on the second step but did not improve the prediction of combination memory performance R2 change = .00 (F< 1). However, on the third step, the main effect of age was added and significantly improved the prediction of combination memory performance, R =.790, R2change = .039, F(2, 74) = 3.85, p < .05. We found significant beta coefficients for the comparison of combination memory performance between third-

Lorsbach & Reimer

323

grade students (M = 1.77) and sixth-grade students (M = 2.62), unstandardized B =.132, 03= .203, t(79) = 2.28, p < .05, and for comparison of combination memory performance of third- and sixth-grade students (M = 2.20) versus college students (M = 3.20), unstandardized B = .182, 3= .161, t(79) = 2.04, p < .05. Therefore, the results of this regression analysis indicated that age was a significant predictor of combination memory performance, even after we accounted for the effects of object and location memory performance and their interaction. These results suggest that the age-related changes found in memory performance for feature conjunctions are not solely the function of memory for the component features. In addition to measures of recognition accuracy, the time required to make a decision provided a useful index of the relative difficulty of feature memory and feature binding. Because younger children perform various cognitive processes more slowly than do older children and adults (e.g., Kail, 1991), we wondered whether the difficulties of children in the feature-only condition and the combination condition would be reflected in their response latencies. To examine this question, we analyzed response latencies and followed the procedure used by K. J. Mitchell, Johnson, Raye, Mather, and D'Esposito (2000). We calculated a difference score for each participant in each memory condition: median reaction time on lures minus median reaction time on targets. We designed these scores to control for differences in base line reaction times by "anchoring participants' lure performance to their own target performance" (K. J. Mitchell, Johnson, Raye, Mather, & D'Esposito, p. 532). The analysis of the lure-target difference scores revealed that only the effect of memory condition was significant, F(2, 154) = 20.108, MSE = 40048.102, p < .0001. Participants were slower on lures on targets in the combination memory condition (M = 183 ms) than on either the object (M = - 6 ms) or the location memory (M = 32 ms) condition, whereas the latter two memory conditions did not differ from one another. These results are not consistent with the position that children experience greater difficulty than do young adults when they try to remember associations between features compared with single features. That all participants were slower to respond in the combination condition than in the feature-only condition provides support for the observation that feature binding is more effortful than feature memory (Reinitz et al., 1994; K. J. Mitchell, Johnson, Raye, Mather, & D'Esposito). Discussion' In the present study, we sought to determine whether developmental differences existed in the ability to bind features together in a working memory task. Analysis of recognition memory performance indicated that, although sixth-grade students were comparable to college students in memory for objects, the sixthgrade students experienced greater difficulty when attempting to remember locations and even greater difficulty when attempting to remember objects and their

I

.324

The Journal of Genetic Psychology

locations, in the combination condition. Third-grade students performed worse than did sixth-grade students in their memory for individual features (objects or their locations), as well as in their memory for the combination of those features. Most important, although memory in the combination condition is correlated with feature memory, developmental differences in the combination condition remain after variance associated with feature memory is removed statistically. Taken together, these results suggest that developmental differences exist in the binding of featural information. One may question whether we obtained the results of the present study because of the failure of children, particularly third-grade students, to understand the procedures associated with each memory task.-Aside from the fact that the tasks were simple and uncomplicated, two major factors would seem to rule out this possibility. First, we did not present test trials until participants demonstrated their understanding of a given memory task on the practice trials. Second, performance on the tasks themselves further indicated that children understood the tasks. If children failed to understand the tasks and to follow instructions, their performance would have been at chance levels. An examination of the results indicates that both groups of children performed at levels that were well above chance. Researchers have focused theoretical accounts of binding on those processes that occur during initial encoding and consolidation. For example, Reinitz and colleagues (e.g., Hannigan & Reinitz, 2000; Reinitz et al., 1994) proposed that individual stimulus features and relational information (global stimulus features) are each encoded separately. The processing of relational information places greater demhnds on processing resources and, consequently, such information is more apt to be forgotten than individual features. Retrieval is presumed to involve a conjunction process whereby stored relational information is used to recombine stored features. In the absence of relational information, conjunction errors occur at the time of retrieval. Thus, one might argue that the binding difficulties of children, who were observed in the combination condition, were the result of a failure to encode relational information. If this interpretation is accurate, the age differences that we observed iný the Combination condition would suggest that there are age-related improvements in the ability to process relational information during a working memory task. Johnson and colleagues (e.g., Johnson, 1992, 1997; Johnson & Chalfonte, 1994; Chalfonte et al., 1996) similarly emphasized the role of encoding processes by proposing that featural binding involves the use of reflective processes. As depicted in the Multiple-Entry, Modular Memory System framework (MEM; Johnson, 1983), reflective processes essentially enable one to "sustain, organize, and revive information" (Johnson & Hirst, 1993, p. 245). The reflective system within MEM includes a number of component processes, including, "noting relations, shifting attention to something potentially more useful, refreshing information so that it remains active and one can easily shift back to it, and reactivating information that has dropped out of consciousness" (Johnson & Hirst, p. 245).

Lorsbach &Reimer

325

Each of these component processes is considered to affect binding. For example, binding is presumably more apt to occur if one notes relations between features (e.g., objects and locations) or refreshes and reactivates two features that cooccurred in a recent event (e.g., object and its location). One might argue that the developmental differences in the combination condition of the present study were caused, in part, by the inefficient use of one or more of these reflective processes. The binding deficits in children might be interpreted in terms of an underlying deficit in reflective processes (Johnson, 1992), particularly those that are involved in the reactivation of information that has left consciousness (Johnson & Hirst). Reactivation is presumed to be critical for memorial binding. By bringing inactive information back to mind, "reactivation acts as an internally generated repetition of the information" and serves "to promote opportunities for binding and to strengthen existing relations" (Chalfonte & Johnson, 1996, p. 214). In the present study, we provided evidence for developmental differences in feature binding. Although there were developmental differences in featural binding, the design of the present study does not allow one to identify the locus of these binding problems. Any observed difficulties with binding in the present paradigm may have been the result of processing differences during (a) the acquisition phase, (b) the retention interval, (c) the test, or (d) some combination of these (K. J. Mitchell, Johnson, Raye, Mather, & D'Esposito, 2000). In the future, researchers will need to examine why there are developmental differences in remembering combinations of stimulus features. REFERENCES Bower, G. H. (1967). A multicomponent theory of the memory trace. In K. W. Spence & J. T. Spence (Eds.), The psychology of learning and motivation:Advances in research and theory (Vol. 1). NewYork: Academic Press. Case, R. (1985). Intellectual development: Birth to adulthood.NewYork: Academic Press. Case, R., Kurland, M., & Goldberg, J. (1982). Operational efficiency and the growth of short-term memory span. Journal of Experimental ChildPsychology, 33, 386-404. Cedrus Corporation. (1989-1999). SuperLab General Purpose Psychology Testing Package: Pro 1.75 [computer software]. Phoenix, AZ: Author. Cerella, J. (1985). Information processing rates in the elderly. PsychologicalBulletin, 98, 67-83. Chalfonte, B. L., & Johnson, M. K. (1996). Feature memory and binding in young and older adults. Memory & Cognition, 24, 403-416. Chalfonte, B. L., Verfaellie, M., Johnson, M. K., & Reiss, L. (1996). Spatial location memory in amnesia: Binding item and location information under incidental and intentional encoding conditions. Memory, 4, 591-614. Cohen, N. J., & Eichenbaum, H. (1993). Memory, amnesia, and the hippocampalsystem. Cambridge, MA: MIT Press. Cycowicz, Y. M., Friedman, D., & Rothstein, M. (1997). Picture naming by young children: Norms for name agreement, familiarity, and visual complexity. Journalof Experimental ChildPsychology, 65, 171-237. Dempster, F. N. (1992). The rise and fall of the inhibitory mechanism: Toward a unified theory of cognitive development and aging. Developmental Review, 12, 45-75.

326

The Journalof Genetic Psychology

Eichenbaum, H., & Bunsey, M. (1995). On the binding of associations in memory: Clues from studies on the role of the hippocampus region in paired-associate learning. Current Directions in PsychologicalScience, 4, 19-23. Hannigan, S. L., & Reinitz, M. T. (2000). Influences of temporal factors on memory conjunction errors. Applied Cognitive Psychology, 14, 309-321. Hasher, L., & Zacks, R. T. (1988). Working memory, comprehension, and aging: A review and a new view. In G. H.. Bower (Ed.), The psychology of learning and motivation: Advances in. research and theory (Vol. 22, pp. 193-225). San Diego, CA: Academic Press. Hautus, M. J. (1995). Corrections for extreme proportions and their biasing effects on estimated values of d'. BehaviorResearch Methods, Instruments, & Computers, 27, 46-51. Huttenlocher, P. R. (1990). Morphometric study of human cerebral cortex development. Neuropsychologia,28, 517-527. Johnson, M. K. (1983). A multiple-entry, modular memory system. In G. H. Bower (Ed.), T7e psychology of learning and motivation:Advances in research and theory (Vol. 17, pp. 81-123). New York: Academic Press. Johnson, M. K. (1992). MEM: Mechanisms of recollection. Journalof Cognitive Neuroscience, 4, 268-280. Johnson, M. K. (1997). Identifying the origin of mental experience. In M. S. Myslobodsky (Ed.), The mythomanias: The natureof deception and self-deception (pp. 133-180). Mahwah, NJ: Erlbaum. Johnson, M. K. (1999). Developmental neuroscience. In M. H. Bornstein & M. E. Lamb (Eds.), Developmentalpsychology: An advanced textbook (pp. 199-230). Mahwah, NJ: Erlbaum. Johnson, M. K., & Chalfonte, B. L. (1994). Binding complex memories: The role of reactivation and the hippocampus. In D. L. Schacter & E. Tulving (Eds.), Memory systems (pp. 311-350). Cambridge, MA: MIT Press. Johnson, M. K., & Hirst, W. (1993). MEM: Memory subsystems as processes. In A. F. Collins, S. E. Gathercole, M. A. Conway, & P. E. Morris (Eds.), Theories of memory (pp. 241-286). East Sussex, England: Erlbaum. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1996). Source monitoring. Psychological Bulletin, 114, 3-28. Kail, R. V. (1991). Development of processing speed in childhood and adolescence. InH. W. Reese (Ed.), Advances in child development and behavior (Vol. 23, pp. 151-185). .San Diego, CA: Academic Press. Kail, R. V., & Siegel, A. W. (1977). Sex differences in retention of verbal and spatial characteristics of stimuli. Journalof Experimental ChildPsychology, 23, 341-347. Kroll, N. E. A., Knight, R. T., Metcalfe, J., Wolf, E. S., & Tulving, E. (1996). Cohesion failure'is a source of memory illusions. JournalofMemor' andLanguage,35, 176-196. McClelland, J. L., & Rumelhart, D. E. (1985). Distributed memory and the representation of general and specific information. Journalof ExperimentalPsychology: General,114, 159-188. Metcalfe, J., Mencl, W. E., & Cottrell, G. W. (1994). Cognitive binding.' In D. L. Schacter & E. Tulving (Eds.), Memory systems (pp. 369-394). Cambridge, MA: MIT Press. Mitchell, D. B. (1991). Implicit and explicit memory for pictures: Multiple views across the lifespan. In P. Graf & M. E. J. Masson (Eds.), Inplicitmemory (pp. 171-190). Hillsdale, NJ: Erlbaum. Mitchell, K. J., Johnson, M. K., Raye, C. L., & D'Esposito, M. D. (2000). MRI evidence of age-related hippocampal dysfunction in feature binding in working memory. Cognitive Brain Research, 10, .197-206. Mitchell, K. J., Johnson, M. K., Raye, C. L., Mather, M., & D'Esposit6, M. D. (2000).

Lorsbach & Reimer

327

Aging and reflective processes of working memory: Binding and test load deficits. Psychology andAging, 15, 427-541. Parkin, A. J. (1991). Implicit memory across the lifespan. In P. Graf & M. E. J. Masson (Eds.), Implicit memory (pp. 191-206). Hillsdale, NJ: Erlbaum. Reinitz, M. T., Morrissey, J., & Demb, J. (1994). Role of attention in face encoding. Journal of Experimental Psychology: Learning,Memory and Cognition, 20, 161-168. Schacter, D. L., Norman, K.A., & Koustaal, W. (1998). The cognitive neuroscience of constructive memory. Annual Review of Psychology, 49, 289-318. Siemens, L., Guttentag, R. E., & McIntyre, M. (1989). Age differences in memory for item-identity and occupied-location information. American Journal ofPsychology, 102, 53-68. Snodgrass, J. G., & Corwin, J. (1988). Pragmatics of measuring recognition memory: Applications to dementia and amnesia. Journalof Experimental Psychology: General, 117, 34-50. Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174-215. Stanislaw, H., & Todorov, N. (1999). Calculation of signal detection theory measures. BehaviorResearch Methods, Instruments, & Computers, 31, 137-149. Underwood, B. J. (1969). Attributes of memory. PsychologicalReview, 76, 559-573. Yakovlev, P. I., & LeCours, A. R. (1967). The myelogenetic cycles of regional maturation of the brain. In A. Minkowski (Ed.), Regional development of the brain in early life (pp. 3-70). Oxford, England: Blackwell. Zechmeister, E. B, & McKillip, J. (1972). Recall of place on the page. Journalof Educational Psychology, 63, 446-453.

Received May 5, 2005

COPYRIGHT INFORMATION

TITLE: Feature Binding in Children and Young Adults SOURCE: J Genet Psychol 166 no3 S 2005 WN: 0524402348008 The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited. To contact the publisher: http://www.heldref.org/

Copyright 1982-2005 The H.W. Wilson Company.

All rights reserved.