Figure 1 - uploaded by Kristy Marie Snyder
Content may be subject to copyright.
Source publication
We address the problem of serial order in skilled typing, asking whether typists represent the identity and order of the keystrokes they type jointly by linking successive keystrokes into a chained sequence, or separately by associating keystrokes with position codes. In 4 experiments, typists prepared to type a prime word and were probed to type a...
Contexts in source publication
Context 1
... target word and probed with the target word ( Figure 1, Panel C). In the unrelated prime condition, typists were primed with a randomly selected word that was not the target word or an anagram of the target word, and then probed with a target word (Figure 1, Panel D). Each of the 194 words served as a prime four times, once per condition, resulting in a total of 776 trials. Typists were instructed to pay attention to the prime word and to prepare to type it as soon as the go stimulus appeared. If a word was displayed instead of the go stimulus, typists were to type that word as quickly and accurately as possible. The backspace key was disabled, so typists were not able to correct their responses. Typists pressed the spacebar to move on to the next trial. Once typists finished the experiment, they completed the typing test. We calculated mean RT from correct trials. We excluded RTs that were more than 2.5 standard deviations from the mean (Van Selst & Jolicoeur, 1994). This excluded 2.6% of the data. We also calculated mean error rates (i.e., percentage of trials in which at least one typing error was committed) for each condition for each typist. Mean RTs across typists are presented in Figure 2. We conducted one-way analyses of variance (ANOVAs) on the RT and error rates. The summary tables for the ANOVAs are presented in Table 1. ANOVA revealed that RT differed significantly between the conditions. To determine which differences were significant, we calculated Fisher’s least significant difference (LSD), which was 60 ms for p Ͻ .05. We also calculated the Bonferroni-corrected minimum mean difference, which was 83 ms for the adjusted alpha level of .008 per test (.05/6). Using either criterion, RT did not differ significantly between go ( M ϭ 558 ms) and target prime trials ( M ϭ 521 ms), which suggests that typists were prepared to type the prime. RT in unrelated prime trials ( M ϭ 692 ms) was significantly longer than both go and target prime trials. RT in anagram prime trials ( M ϭ 706 ms) was significantly longer than in target prime trials. RT did not differ significantly between anagram and unrelated trials ( M ϭ 692 ms). These findings indicate that it takes the motor system approximately the same amount of time to reorder a set of previously activated keystrokes as it does to activate and order a new set of keystrokes. These results are consistent with the notion that item and order information are represented jointly, not separately, in skilled typewriting. There were no significant differences in error rates between the conditions (go: M ϭ 10.1%; target: M ϭ 9.2%; anagram: M ϭ 10.3%; unrelated: M ϭ 10.3%). The results of Experiment 1 indicated that key identities were not primed separately from their order. The findings were more consistent with chaining theories than position coding theories, but they do not rule out position coding theories entirely. Although keystroke identities were the same in target and anagram primes, the keystrokes that were associated with position codes differed. If typists represent serial order by position coding, priming may be produced only when keystroke identities are associated with the same position code in the prime and the target. In Experiment 2, we tested whether a prime would facilitate RT if two of the keystroke identities in the prime and target were associated with the same position codes, but occurred in different sequences. We compared primes in which keystroke identities were the same as the target in the first two positions but differed in the last two positions (SSDD trials), primes in which keystroke identities were the same as the target in the last two positions but differed in the first two positions (DDSS trials), primes in which keystroke identities were the same as the target in all four positions (SSSS trials), and primes in which the keystroke identities differed from the target in all four positions (DDDD trials), in which S and D indicate whether the keystroke identity associated with a position was the same (S) or different (D) in the prime and target. If serial order is represented by position coding, RT should be faster for targets that follow both SSDD and DDSS primes than for targets that follow DDDD primes. When targets follow SSDD and DDSS primes, the associations between two keystrokes and two positions are the same in the prime and the target. The activation in these positions should reduce the amount of time needed to prepare a motor program for the target. If serial order is represented by chaining, RT should be faster for targets that follow SSDD primes than for targets that follow DDDD primes, but RT should not be faster targets that follow DDSS primes than for targets that follow DDDD primes. Chaining theories assume that motor programs are prepared from the beginning of the sequence. In SSDD trials, the first two links in the chain are primed, which should reduce the amount of time needed to prepare them. In DDSS trials, the last two keystrokes are primed, but they ...
Context 2
... target word and probed with the target word ( Figure 1, Panel C). In the unrelated prime condition, typists were primed with a randomly selected word that was not the target word or an anagram of the target word, and then probed with a target word (Figure 1, Panel D). Each of the 194 words served as a prime four times, once per condition, resulting in a total of 776 trials. Typists were instructed to pay attention to the prime word and to prepare to type it as soon as the go stimulus appeared. If a word was displayed instead of the go stimulus, typists were to type that word as quickly and accurately as possible. The backspace key was disabled, so typists were not able to correct their responses. Typists pressed the spacebar to move on to the next trial. Once typists finished the experiment, they completed the typing test. We calculated mean RT from correct trials. We excluded RTs that were more than 2.5 standard deviations from the mean (Van Selst & Jolicoeur, 1994). This excluded 2.6% of the data. We also calculated mean error rates (i.e., percentage of trials in which at least one typing error was committed) for each condition for each typist. Mean RTs across typists are presented in Figure 2. We conducted one-way analyses of variance (ANOVAs) on the RT and error rates. The summary tables for the ANOVAs are presented in Table 1. ANOVA revealed that RT differed significantly between the conditions. To determine which differences were significant, we calculated Fisher’s least significant difference (LSD), which was 60 ms for p Ͻ .05. We also calculated the Bonferroni-corrected minimum mean difference, which was 83 ms for the adjusted alpha level of .008 per test (.05/6). Using either criterion, RT did not differ significantly between go ( M ϭ 558 ms) and target prime trials ( M ϭ 521 ms), which suggests that typists were prepared to type the prime. RT in unrelated prime trials ( M ϭ 692 ms) was significantly longer than both go and target prime trials. RT in anagram prime trials ( M ϭ 706 ms) was significantly longer than in target prime trials. RT did not differ significantly between anagram and unrelated trials ( M ϭ 692 ms). These findings indicate that it takes the motor system approximately the same amount of time to reorder a set of previously activated keystrokes as it does to activate and order a new set of keystrokes. These results are consistent with the notion that item and order information are represented jointly, not separately, in skilled typewriting. There were no significant differences in error rates between the conditions (go: M ϭ 10.1%; target: M ϭ 9.2%; anagram: M ϭ 10.3%; unrelated: M ϭ 10.3%). The results of Experiment 1 indicated that key identities were not primed separately from their order. The findings were more consistent with chaining theories than position coding theories, but they do not rule out position coding theories entirely. Although keystroke identities were the same in target and anagram primes, the keystrokes that were associated with position codes differed. If typists represent serial order by position coding, priming may be produced only when keystroke identities are associated with the same position code in the prime and the target. In Experiment 2, we tested whether a prime would facilitate RT if two of the keystroke identities in the prime and target were associated with the same position codes, but occurred in different sequences. We compared primes in which keystroke identities were the same as the target in the first two positions but differed in the last two positions (SSDD trials), primes in which keystroke identities were the same as the target in the last two positions but differed in the first two positions (DDSS trials), primes in which keystroke identities were the same as the target in all four positions (SSSS trials), and primes in which the keystroke identities differed from the target in all four positions (DDDD trials), in which S and D indicate whether the keystroke identity associated with a position was the same (S) or different (D) in the prime and target. If serial order is represented by position coding, RT should be faster for targets that follow both SSDD and DDSS primes than for targets that follow DDDD primes. When targets follow SSDD and DDSS primes, the associations between two keystrokes and two positions are the same in the prime and the target. The activation in these positions should reduce the amount of time needed to prepare a motor program for the target. If serial order is represented by chaining, RT should be faster for targets that follow SSDD primes than for targets that follow DDDD primes, but RT should not be faster targets that follow DDSS primes than for targets that follow DDDD primes. Chaining theories assume that motor programs are prepared from the beginning of the sequence. In SSDD trials, the first two links in the chain are primed, which should reduce the amount of time needed to prepare them. In DDSS trials, the last two keystrokes are primed, but they follow two unprimed keystrokes. A new chain would need to be established from the beginning of the sequence, so the advantage of priming would be lost. It is possible that the priming effects on RT ...
Context 3
... separately from their order. Position coding theories of serial order suggest that item and order information are represented separately, so priming a sequence of keystrokes should activate the identities of all the keystrokes separately from their order. Chaining theories suggest that item and order information are represented jointly, so priming a sequence of keystrokes should activate keystroke identities only in the sequence that was primed. On each trial, typists were presented with a prime word and were told to prepare to type it (see Figure 1). When typists prepare to type a word, the motor system activates the relevant item and order information and maintains this information in a motor program. To ensure that the typists prepared to type the prime, the prime was followed by a go signal (i.e., ءءءءء ) that prompted the typists to type the prime in 25% of the trials. In the remaining trials, the prime was followed by a target that was identical to the prime, an anagram of the prime, or a word that was unrelated to the prime. When the target is identical to the prime, the motor system only needs to execute the motor program that was prepared for the prime. When the target is unrelated to the prime, the motor program that was prepared for the prime is no longer appropriate, so the motor system needs to generate a new motor program by activating and ordering a new set of keystrokes. As a result, RT should be shorter when targets follow identical primes than when they follow unrelated primes. When the target is an anagram of the prime, the motor program that was prepared for the prime activated all of the keystrokes necessary to type the target, but in an inappropriate order. Thus, the motor program needs to be changed. The cognitive system may do this by creating a new program or revising the existing program. In either case, the residual activation of the shared keystrokes will affect RT differently depending on whether item and order information are represented jointly or separately. If item and order are represented jointly, as serial chaining theories suggest, the motor system would need to activate a chain of keystrokes that are linked in a specific sequence, just as it would when targets follow unrelated primes. Thus, RT should be as long when targets follow anagram primes as when targets follow unrelated primes. If item and order are represented separately, as position coding theories suggest, the motor system would need to associate the active keystrokes with different position codes. The activation of the keystrokes may decrease the amount of time it takes to create a motor program for the target, so RT may be shorter when targets follow anagram primes than when they follow unrelated primes. However, the activated keystrokes may have to be dissociated from the positions codes they were associated with and reassigned to new position codes. That may increase the time it takes the motor system to create a motor program for the target, so RT may be longer when targets follow anagram primes than when they follow unrelated primes (Neill & Mathis, 1998; Neill, Valdes, Terry, & Gorfein, 1992). In either case, the motor system has to associate the keystrokes with the appropriate position codes, so RT should be longer when targets follow anagram primes than when they follow identical primes. Subjects. We recruited 17 typists who had formal training in touch typing and the self-reported ability to type 40 WPM. We did not use the data from one typist who did not follow task instructions. We verified their typing skill with a typing test (for details, see WPM Logan & Zbrodoff, 1998). Their average typing speed was 72.2 WPM (range ϭ 43.6 –121.5 WPM) and their mean accuracy was 93.4% (range ϭ 82.9%–100%). They received course credit or $12 for 60 min of participation. Apparatus and materials. We compiled a pool of 194 five- letter words from the MRC Psycholinguistic Database (Wilson, 1987). The mean word frequency per million words was 52.0 (range ϭ .01–1139.2), as verified by the Corpus of Contemporary American English (Davies, 2008). Each word was an anagram of another word in the list (see Appendix A). No anagrams shared first letters. The experiment took place on a personal computer programmed in LIVECODE () using a 15-in. SVGA monitor. Typists sat about 57 cm from the monitor. Responses were registered on a standard QWERTY keyboard. The program black- ened the screen and displayed a 24.1 cm ϫ 19.7 cm gray window. The prime word was displayed 5.1 cm from the top of the window in black 40-point Helvetica font. The probe was presented 6.4 cm from the top of the window in the same font. Typists’ responses were echoed 3.8 cm below the probe. Procedure. At the beginning of each trial, a prime word was displayed for 250 ms. It was subsequently removed. After a 500-ms blank interval, the probe was displayed. The probe was either a go stimulus (i.e., ءءءءء ) or a target word. There were four conditions (see Figure 1). In the go condition, typists were primed with a word and probed with the go stimulus (Figure 1, Panel A). In the target prime condition, typists were primed with the target word and probed with the target word (Figure 1, Panel B). In the anagram prime condition, typists were primed with an anagram ...
Context 4
... information are represented separately, so priming a sequence of keystrokes should activate the identities of all the keystrokes separately from their order. Chaining theories suggest that item and order information are represented jointly, so priming a sequence of keystrokes should activate keystroke identities only in the sequence that was primed. On each trial, typists were presented with a prime word and were told to prepare to type it (see Figure 1). When typists prepare to type a word, the motor system activates the relevant item and order information and maintains this information in a motor program. To ensure that the typists prepared to type the prime, the prime was followed by a go signal (i.e., ءءءءء ) that prompted the typists to type the prime in 25% of the trials. In the remaining trials, the prime was followed by a target that was identical to the prime, an anagram of the prime, or a word that was unrelated to the prime. When the target is identical to the prime, the motor system only needs to execute the motor program that was prepared for the prime. When the target is unrelated to the prime, the motor program that was prepared for the prime is no longer appropriate, so the motor system needs to generate a new motor program by activating and ordering a new set of keystrokes. As a result, RT should be shorter when targets follow identical primes than when they follow unrelated primes. When the target is an anagram of the prime, the motor program that was prepared for the prime activated all of the keystrokes necessary to type the target, but in an inappropriate order. Thus, the motor program needs to be changed. The cognitive system may do this by creating a new program or revising the existing program. In either case, the residual activation of the shared keystrokes will affect RT differently depending on whether item and order information are represented jointly or separately. If item and order are represented jointly, as serial chaining theories suggest, the motor system would need to activate a chain of keystrokes that are linked in a specific sequence, just as it would when targets follow unrelated primes. Thus, RT should be as long when targets follow anagram primes as when targets follow unrelated primes. If item and order are represented separately, as position coding theories suggest, the motor system would need to associate the active keystrokes with different position codes. The activation of the keystrokes may decrease the amount of time it takes to create a motor program for the target, so RT may be shorter when targets follow anagram primes than when they follow unrelated primes. However, the activated keystrokes may have to be dissociated from the positions codes they were associated with and reassigned to new position codes. That may increase the time it takes the motor system to create a motor program for the target, so RT may be longer when targets follow anagram primes than when they follow unrelated primes (Neill & Mathis, 1998; Neill, Valdes, Terry, & Gorfein, 1992). In either case, the motor system has to associate the keystrokes with the appropriate position codes, so RT should be longer when targets follow anagram primes than when they follow identical primes. Subjects. We recruited 17 typists who had formal training in touch typing and the self-reported ability to type 40 WPM. We did not use the data from one typist who did not follow task instructions. We verified their typing skill with a typing test (for details, see WPM Logan & Zbrodoff, 1998). Their average typing speed was 72.2 WPM (range ϭ 43.6 –121.5 WPM) and their mean accuracy was 93.4% (range ϭ 82.9%–100%). They received course credit or $12 for 60 min of participation. Apparatus and materials. We compiled a pool of 194 five- letter words from the MRC Psycholinguistic Database (Wilson, 1987). The mean word frequency per million words was 52.0 (range ϭ .01–1139.2), as verified by the Corpus of Contemporary American English (Davies, 2008). Each word was an anagram of another word in the list (see Appendix A). No anagrams shared first letters. The experiment took place on a personal computer programmed in LIVECODE () using a 15-in. SVGA monitor. Typists sat about 57 cm from the monitor. Responses were registered on a standard QWERTY keyboard. The program black- ened the screen and displayed a 24.1 cm ϫ 19.7 cm gray window. The prime word was displayed 5.1 cm from the top of the window in black 40-point Helvetica font. The probe was presented 6.4 cm from the top of the window in the same font. Typists’ responses were echoed 3.8 cm below the probe. Procedure. At the beginning of each trial, a prime word was displayed for 250 ms. It was subsequently removed. After a 500-ms blank interval, the probe was displayed. The probe was either a go stimulus (i.e., ءءءءء ) or a target word. There were four conditions (see Figure 1). In the go condition, typists were primed with a word and probed with the go stimulus (Figure 1, Panel A). In the target prime condition, typists were primed with the target word and probed with the target word (Figure 1, Panel B). In the anagram prime condition, typists were primed with an anagram ...
Context 5
... cued. The system does not have to wait for the item associated with the position code to be accessed before cuing the next position code. Theories of skilled typing assume that reaction time (RT) for the first keystroke reflects the time it takes to encode the material to be typed, prepare a motor program to type it, and implement the first step of the program (i.e., execute the first keystroke). Interkeystroke latencies reflect the time it takes to implement each successive step of the program (Crump & Logan, 2010; Logan & Crump, 2011; Salthouse, 1986). Motor programs are representations that specify the movements to be executed to achieve a goal and their order of execution (Keele, Cohen, & Ivry, 1990; Miller, Galanter, & Pribram, 1960). The assumption that motor programs specify all of the movements before the first movement is executed is sup- ported by at least three lines of evidence: The time it takes to begin an action increases with the complexity of the action, consecutive movements are frequently coarticulated, and initial movements frequently respect end-state rather than beginning-state comfort (Keele, 1968; Rosenbaum, Cohen, Jax, Weiss, & van der Wel, 2007; Rosenbaum, Engelbrecht, Bushe, & Loukopoulos, 1993; Rosenbaum, Hindorff, & Munro, 1987). Crump and Logan (2010) used a priming task to show that motor programs specify the identity and the order of the keystrokes typists type. They presented a prime word followed by a probe—a single letter or a word—which typists had to type as quickly and accurately as possible. RT was faster when probe letters were from the primed word than when they were not, suggesting that the prime word activated all of the constituent letters in parallel. We adapted Crump and Logan’s (2010) priming paradigm to investigate the problem of serial order in skilled typing. Skilled typists were shown a prime word and were then probed to type a target word. We varied the overlap between the identity and the order of letters in the prime and target. We focused primarily on RT, defined as the interval between the onset of the probe and the execution of the first keystroke, because it reflects the time it takes to create a motor program (plus the time for word encoding and executing the first keystroke). Following Crump and Logan, we assumed that the prime would activate the motor program required to type it, and that would affect the time required to create and implement the motor program for the target. The greater the overlap between prime and target, the less time required to program the target, and the shorter the RT. We conducted four experiments using the priming paradigm to distinguish between chaining and position coding theories of serial order in skilled typing. First, we asked whether keystroke identities could be primed separately from the position they occupy in a sequence. Experiment 1 compared the RT of probed target words that were preceded by identity, anagram, and unrelated primes. The results indicated that keystroke identities could not be primed separately from their order. Second, we asked whether keystroke identities could be primed in the correct order but out of sequence. Experiment 2 compared priming the first two keystrokes with priming the last two keystrokes. The results indicated that keystroke identities could be primed when they occur in the correct order at the beginning of the sequence, but not when they occur at the end of the sequence. Experiment 3 compared priming the first and last keystrokes with priming the middle two keystrokes. The results indicated that priming the first and last keystrokes produced an RT advantage, but priming the middle two keystrokes did not. The RT advantage observed when the first two keystrokes were primed in Experiment 2 was more pronounced than the RT advantage observed when the first and last keystrokes were primed in Experiment 3. Experiment 4 asked whether priming has a graded influence that depends on the number of keystrokes that are primed in sequence. We compared RT for probed target words that were preceded by primes that shared the first one, two, three, or four keystrokes with the target. The results indicated that priming increased with the number of keystrokes that were primed in sequence and was graded across the keystroke sequence, with less priming in later positions. The purpose of Experiment 1 was to determine whether keystroke identities can be primed separately from their order. Position coding theories of serial order suggest that item and order information are represented separately, so priming a sequence of keystrokes should activate the identities of all the keystrokes separately from their order. Chaining theories suggest that item and order information are represented jointly, so priming a sequence of keystrokes should activate keystroke identities only in the sequence that was primed. On each trial, typists were presented with a prime word and were told to prepare to type it (see Figure 1). When typists prepare to type a word, the motor system activates the relevant item and order information and maintains this information in a motor program. To ensure that the typists prepared to type the prime, the prime was followed by a go signal (i.e., ءءءءء ) that prompted the typists to type the prime in 25% of the trials. In the remaining trials, the prime was followed by a target that was identical to the prime, an anagram of the prime, or a word that was unrelated to the prime. When the target is identical to the prime, the motor system only needs to execute the motor program that was prepared for the prime. When the target is unrelated to the prime, the motor program that was prepared for the prime is no longer appropriate, so the motor system needs to generate a new motor program by activating and ordering a new set of keystrokes. As a result, RT should be shorter when targets follow identical primes than when they follow unrelated primes. When the target is an anagram of the prime, the motor program that was prepared for the prime activated all of the keystrokes necessary to type the target, but in an inappropriate order. Thus, the motor program needs to be changed. The cognitive system may do this by creating a new program or revising the existing program. In either case, the residual activation of the shared keystrokes will affect RT differently depending on whether item and order information are represented jointly or separately. If item and order are represented jointly, as serial chaining theories suggest, the motor system would need to activate a chain of keystrokes that are linked in a specific sequence, just as it would when targets follow unrelated primes. Thus, RT should be as long when targets follow anagram primes as when targets follow unrelated primes. If item and order are represented separately, as position coding theories suggest, the motor system would need to associate the active keystrokes with different position codes. The activation of the keystrokes may decrease the amount of time it takes to create a motor program for the target, so RT may be shorter when targets follow anagram primes than when they follow unrelated primes. However, the activated keystrokes may have to be dissociated from the positions codes they were associated with and reassigned to new position codes. That may increase the time it takes the motor system to create a motor program for the target, so RT may be longer when targets follow anagram primes than when they follow unrelated primes (Neill & Mathis, 1998; Neill, Valdes, Terry, & Gorfein, 1992). In either case, the motor system has to associate the keystrokes with the appropriate position codes, so RT should be longer when targets follow anagram primes than when they follow identical primes. Subjects. We recruited 17 typists who had formal training in touch typing and the self-reported ability to type 40 WPM. We did not use the data from one typist who did not follow task instructions. We verified their typing skill with a typing test (for details, see WPM Logan & Zbrodoff, 1998). Their average typing speed was 72.2 WPM (range ϭ 43.6 –121.5 WPM) and their mean accuracy was 93.4% (range ϭ 82.9%–100%). They received course credit or $12 for 60 min of participation. Apparatus and materials. We compiled a pool of 194 five- letter words from the MRC Psycholinguistic Database (Wilson, 1987). The mean word frequency per million words was 52.0 (range ϭ .01–1139.2), as verified by the Corpus of Contemporary American English (Davies, 2008). Each word was an anagram of another word in the list (see Appendix A). No anagrams shared first letters. The experiment took place on a personal computer programmed in LIVECODE () using a 15-in. SVGA monitor. Typists sat about 57 cm from the monitor. Responses were registered on a standard QWERTY keyboard. The program black- ened the screen and displayed a 24.1 cm ϫ 19.7 cm gray window. The prime word was displayed 5.1 cm from the top of the window in black 40-point Helvetica font. The probe was presented 6.4 cm from the top of the window in the same font. Typists’ responses were echoed 3.8 cm below the probe. Procedure. At the beginning of each trial, a prime word was displayed for 250 ms. It was subsequently removed. After a 500-ms blank interval, the probe was displayed. The probe was either a go stimulus (i.e., ءءءءء ) or a target word. There were four conditions (see Figure 1). In the go condition, typists were primed with a word and probed with the go stimulus (Figure 1, Panel A). In the target prime condition, typists were primed with the target word and probed with the target word (Figure 1, Panel B). In the anagram prime condition, typists were primed with an anagram ...
Context 6
... separately from their order. Chaining theories suggest that item and order information are represented jointly, so priming a sequence of keystrokes should activate keystroke identities only in the sequence that was primed. On each trial, typists were presented with a prime word and were told to prepare to type it (see Figure 1). When typists prepare to type a word, the motor system activates the relevant item and order information and maintains this information in a motor program. To ensure that the typists prepared to type the prime, the prime was followed by a go signal (i.e., ءءءءء ) that prompted the typists to type the prime in 25% of the trials. In the remaining trials, the prime was followed by a target that was identical to the prime, an anagram of the prime, or a word that was unrelated to the prime. When the target is identical to the prime, the motor system only needs to execute the motor program that was prepared for the prime. When the target is unrelated to the prime, the motor program that was prepared for the prime is no longer appropriate, so the motor system needs to generate a new motor program by activating and ordering a new set of keystrokes. As a result, RT should be shorter when targets follow identical primes than when they follow unrelated primes. When the target is an anagram of the prime, the motor program that was prepared for the prime activated all of the keystrokes necessary to type the target, but in an inappropriate order. Thus, the motor program needs to be changed. The cognitive system may do this by creating a new program or revising the existing program. In either case, the residual activation of the shared keystrokes will affect RT differently depending on whether item and order information are represented jointly or separately. If item and order are represented jointly, as serial chaining theories suggest, the motor system would need to activate a chain of keystrokes that are linked in a specific sequence, just as it would when targets follow unrelated primes. Thus, RT should be as long when targets follow anagram primes as when targets follow unrelated primes. If item and order are represented separately, as position coding theories suggest, the motor system would need to associate the active keystrokes with different position codes. The activation of the keystrokes may decrease the amount of time it takes to create a motor program for the target, so RT may be shorter when targets follow anagram primes than when they follow unrelated primes. However, the activated keystrokes may have to be dissociated from the positions codes they were associated with and reassigned to new position codes. That may increase the time it takes the motor system to create a motor program for the target, so RT may be longer when targets follow anagram primes than when they follow unrelated primes (Neill & Mathis, 1998; Neill, Valdes, Terry, & Gorfein, 1992). In either case, the motor system has to associate the keystrokes with the appropriate position codes, so RT should be longer when targets follow anagram primes than when they follow identical primes. Subjects. We recruited 17 typists who had formal training in touch typing and the self-reported ability to type 40 WPM. We did not use the data from one typist who did not follow task instructions. We verified their typing skill with a typing test (for details, see WPM Logan & Zbrodoff, 1998). Their average typing speed was 72.2 WPM (range ϭ 43.6 –121.5 WPM) and their mean accuracy was 93.4% (range ϭ 82.9%–100%). They received course credit or $12 for 60 min of participation. Apparatus and materials. We compiled a pool of 194 five- letter words from the MRC Psycholinguistic Database (Wilson, 1987). The mean word frequency per million words was 52.0 (range ϭ .01–1139.2), as verified by the Corpus of Contemporary American English (Davies, 2008). Each word was an anagram of another word in the list (see Appendix A). No anagrams shared first letters. The experiment took place on a personal computer programmed in LIVECODE () using a 15-in. SVGA monitor. Typists sat about 57 cm from the monitor. Responses were registered on a standard QWERTY keyboard. The program black- ened the screen and displayed a 24.1 cm ϫ 19.7 cm gray window. The prime word was displayed 5.1 cm from the top of the window in black 40-point Helvetica font. The probe was presented 6.4 cm from the top of the window in the same font. Typists’ responses were echoed 3.8 cm below the probe. Procedure. At the beginning of each trial, a prime word was displayed for 250 ms. It was subsequently removed. After a 500-ms blank interval, the probe was displayed. The probe was either a go stimulus (i.e., ءءءءء ) or a target word. There were four conditions (see Figure 1). In the go condition, typists were primed with a word and probed with the go stimulus (Figure 1, Panel A). In the target prime condition, typists were primed with the target word and probed with the target word (Figure 1, Panel B). In the anagram prime condition, typists were primed with an anagram ...
Similar publications
This paper deals with the consequences of some manufacturing imprecision related to the use of a fractional-slot concentrated winding (FSCW) in recent permanent-magnet synchronous machine (PMSM) designs. Although these topologies improve the efficiency of motors, they also lead to consider complex models, including undesirable harmonics, which can...
The air-cooled motor controller has been widely used in electric aircraft due to its simple heat dissipation structure and maintenance-free features. Being the main heating component of the electric propulsion system of electric aircraft, the weight and the volume of the air-cooled motor controller need to be strictly controlled. Due to the short p...
Haptic exploration usually involves stereotypical systematic movements that are adapted to the task. Here we tested whether exploration movements are also driven by physical stimulus features. We designed haptic stimuli, whose surface relief varied locally in spatial frequency, height, orientation, and anisotropy. In Experiment 1, participants subs...
Citations
... Although assuming a positional frame is common in models of language production, including orthographic production (e.g., Houghton, 2018), certain findings also indicate a role for chaining mechanisms. For example, Snyder and Logan (2014) reported that primes facilitated target typing the most when primes and targets overlapped in initial sequences, and the greater this overlap, the larger the magnitude of the facilitation. ...
... Finally, we found robust learning in both onset and coda positions, with a larger magnitude of learning in the onset position. While the strength of onset versus coda effects in first-order constraint learning has not been examined before, the asymmetry observed here is compatible with generally stronger effects on initial segments in typing; for example, Snyder and Logan (2014) showed greater facilitation for prime targets overlapping in the initial (e.g., "busy," "burn") than in the final (e.g., "busy," "easy") segments. Experiment 2 investigated the learning of second-order constraints in typing. ...
... Third, there is at least some evidence for chaining as an important mechanism in typing. Snyder and Logan (2014) conducted a series of experiments to test the importance of sequential priming, which directly tests the predictions of chaining models. They found that anagrams (e.g., "ocean") did not prime targets (e.g., "canoe"), showing that the position of primed segments mattered. ...
Possible spoken and written sequences of a language are determined by phonotactic and orthotactic rules, respectively. Adult speakers can learn both simple new phonotactic rules (e.g., “/k/ is always an onset in a syllable”) and more complex second-order rules (e.g., “/k/ is an onset only if the vowel is /æ/, but a coda if the vowel is /ɪ/”). However, the learning timeline for more complex rules is less consistent across populations and languages. In this article, we investigate the learning of parallel orthotactic rules in typing. We first show that adults quickly learn new first-order constraints in typing similar to those in speaking (Experiment 1). Next, we show that they also learn second-order rules, with a timeline similar to learning such phonotactic rules in speaking (Experiment 2). We further find that the second-order constraint is learned for the coda, but not the onset, suggesting that learning new rules of sequencing is carried out by a chaining-type mechanism. Finally, we show that while phonology clearly influences orthography, orthotactic learning cannot be reduced to phonotactic learning (Experiment 3). Collectively, these data support strong similarities between the statistical learning of orthotactic and phonotactic constraints, pointing to the domain generality of the incremental learning principles across different modalities of language production.
... For successful performance of such tasks, the next movement needs to be proactively planned before the previous movement is concluded. Indeed, prior investigations in saccadic eye movements (McPeek et al., 2000;McPeek and Keller, 2002), reading (Rayner, 1998), walking (Patla and Vickers, 2003), typing (Snyder and Logan, 2014), finger movements (Ariani et al., 2021;Ariani et al., 2020;Shahbazi et al., 2024), path tracking (Bashford et al., 2022), target harvesting (Diamond et al., 2017), and reaching (Howard et al., 2015;Säfström et al., 2014;Zimnik and Churchland, 2021) consistently show that sequence production is faster and more efficient when participants have access to information that allows them to plan the future movements. This improvement demonstrates the nervous system's ability to plan future movements while executing the current movement -i.e., to do online planning (Ariani et al., 2021;Ariani et al., 2020;Ariani and Diedrichsen, 2019). ...
Real-world actions often comprise a series of movements that cannot be entirely planned before initiation. When these actions are executed rapidly, the planning of multiple future movements needs to occur simultaneously with the ongoing action. How the brain solves this task remains unknown. Here, we address this question with a new sequential arm reaching paradigm that manipulates how many future reaches are available for planning while controlling execution of the ongoing reach. We show that participants plan at least two future reaches simultaneously with an ongoing reach. Further, the planning processes of the two future reaches are not independent of one another. Evidence that the planning processes interact is twofold. First, correcting for a visual perturbation of the ongoing reach target is slower when more future reaches are planned. Second, the curvature of the current reach is modified based on the next reach only when their planning processes temporally overlap. These interactions between future planning processes may enable smooth production of sequential actions by linking individual segments of a long sequence at the level of motor planning.
... For successful performance of such tasks, the next movement needs to be proactively planned before the previous movement is concluded. Indeed, prior investigations of in saccadic eye movements (McPeek et al., 2000 ;McPeek and Keller, 2002 ), reading (Rayner, 1998 ), walking (Patla and Vickers, 2003 ), typing (Snyder and Logan, 2014 ), finger movements (Ariani et al., 2021(Ariani et al., , 2020Shahbazi et al., 2024 ), path tracking (Bashford et al., 2022 ), target harvesting (Diamond et al., 2017 ), and reaching (Howard et al., 2015 ;Safstrom et al., 2014 ;Zimnik and Churchland, 2021 ) consistently show that sequence production is faster and more efficient when participants have access to information that allows them to plan the future movements. This improvement demonstrates the nervous system's ability to plan future movements while executing the current movement-i.e., to do online planning (Ariani et al., 2021(Ariani et al., , 2020Ariani and Diedrichsen, 2019 ). ...
Real world actions often comprise of a series of movements that cannot be entirely planned before initiation. When these actions are executed rapidly, the planning of multiple future movements needs to occur simultaneously with the ongoing action. How the brain solves this task remains unknown. Here we address this question with a new sequential arm reaching paradigm that manipulates how many future reaches are available for planning while controlling execution of the ongoing reach. We show that participants plan at least two future reaches simultaneously with an ongoing reach. Further, the planning processes of the two future reaches are not independent of one another. Evidence that the planning processes interact is two-fold. First, correcting for a visual perturbation of the ongoing reach target is slower when more future reaches are planned. Second, the curvature of the current reach is modified based on the next reach only when their planning processes temporally overlap. These interactions between future planning processes may enable smooth production of sequential actions by linking individual segments of a long sequence at the level of motor planning.
... For successful performance of such tasks, the next movement needs to be proactively planned before the previous movement is concluded. Indeed, prior investigations of in saccadic eye movements (McPeek et al., 2000;McPeek and Keller, 2002), reading (Rayner, 1998), walking (Patla and Vickers, 2003), typing (Snyder and Logan, 2014), finger movements (Ariani et al., 2021(Ariani et al., , 2020Shahbazi et al., 2023), path tracking (Bashford et al., 2022), target harvesting (Diamond et al., 2017), and reaching (Howard et al., 2015;Safstrom et al., 2014;Zimnik and Churchland, 2021) consistently show that sequence production is faster and more efficient when participants have access to information that allows them to plan the future movements. This improvement demonstrates the nervous system's ability to plan future movements while executing the current movement-i.e., to do online planning (Ariani et al., 2021(Ariani et al., , 2020Ariani and Diedrichsen, 2019). ...
Real world actions often comprise of a series of movements that cannot be entirely planned before initiation. When these actions are executed rapidly, the planning of multiple future movements needs to occur simultaneously with the ongoing action. How the brain solves this task remains unknown. Here we address this question with a new sequential arm reaching paradigm that manipulates how many future reaches are available for planning while controlling execution of the ongoing reach. We show that participants plan at least two future reaches simultaneously with an ongoing reach and that the planning processes of the two future reaches are not independent of one another. Evidence for such interactions is two-fold. First, correcting for a visual perturbation of the ongoing reach target is slower when more future reaches are planned. Second, the curvature of the current reach is modified based on the next reach only when their planning processes temporally overlap. These interactions between future planning processes may enable smooth production of sequential actions by linking individual segments of a long sequence at the level of motor planning.
... For successful performance of such tasks, the next movement needs to be proactively planned before the previous movement is concluded. Indeed, investigations in saccadic eye movement (McPeek et al., 2000;McPeek and Keller, 2002), reading (Rayner, 1998), walking (Patla and Vickers, 2003), typing (Snyder and Logan, 2014), finger movements (Ariani et al., 2021(Ariani et al., , 2020, path tracking (Bashford et al., 2022), and reaching (Safstrom et al., 2014;Zimnik and Churchland, 2021) consistently show that movement production is faster and more efficient when participants have access to information for planning the future movements. This improvement demonstrates the nervous system's ability to plan future movements while executing the current movementi.e., to do online planning (Ariani et al., 2021(Ariani et al., , 2020Ariani and Diedrichsen, 2019). ...
Real world actions often comprise a series of movements that cannot be entirely planned before initiation. When these actions are executed rapidly, planning of future movements needs to occur simultaneously with ongoing execution. However, it remains unknown how the human brain solves this task and whether planning processes of subsequent movements interact. Here we introduce a new sequential reaching paradigm in humans (N=10, 7 sessions each) with a horizon manipulation that allows us to study this interaction by controlling the timing and the overlap of the planning processes for individual movements embedded in the sequence. We show that at least two future reaches are planned simultaneously with the ongoing reach. Two results indicate that these planning processes are not independent of one another. First, correcting an ongoing reach is slower when future movements are planned. Second, the curvature of the current reach is modified based on the next reach only when the planning processes of the two reaches overlap sufficiently. The interactions between future planning processes may enable smooth production of sequential actions.
... With the goal of replicating and extending the results of Experiment 1, we next conducted Experiment 2 as an online study that uses typed responses instead of speech. Typing presents an interesting alternative to speech, because it is slower and more protracted, which could result in different planning strategies (Snyder and Logan, 2014). In our study, the slower time course of typing might affect participants' production decisions, because they have more time to plan upcoming components even after production has begun. ...
... Results from Experiment 2 largely replicated those of Experiment 1: participants were more likely to name the overlapping object first, and this provided a benefit in typing initiation latencies. Despite numerous differences between speech and typing that could result in different strategies for production planning (Snyder and Logan, 2014), we find similar word order biases across modalities, indicating a robust planning bias. The converging findings support our hypothesis that producers can use partial message information to begin planning their responses early, with implications for word order and response times. ...
Language researchers view utterance planning as implicit decision-making: producers must choose the words, sentence structures, and various other linguistic features to communicate their message. To date, much of the research on utterance planning has focused on situations in which the speaker knows the full message to convey. Less is known about circumstances in which speakers begin utterance planning before they are certain about their message. In three picture-naming experiments, we used a novel paradigm to examine how speakers plan utterances before a full message is known. In Experiments 1 and 2, participants viewed displays showing two pairs of objects, followed by a cue to name one pair. In an Overlap condition, one object appeared in both pairs, providing early information about one of the objects to name. In a Different condition, there was no object overlap. Across both spoken and typed responses, participants tended to name the overlapping target first in the Overlap condition, with shorter initiation latencies compared with other utterances. Experiment 3 used a semantically constraining question to provide early information about the upcoming targets, and participants tended to name the more likely target first in their response. These results suggest that in situations of uncertainty, producers choose word orders that allow them to begin early planning. Producers prioritize message components that are certain to be needed and continue planning the rest when more information becomes available. Given similarities to planning strategies for other goal-directed behaviors, we suggest continuity between decision-making processes in language and other cognitive domains.
... We hypothesized that if the pattern of local Position × Type interactions in the manual task was due to the separability between onsets and offsets in our templates, then a language task that introduces such separation in the nonword stimuli would yield a similar pattern of results-local facilitation when onset is repeated. This would suggest that the difference between the patterns seen in the Experiment 1 tasks could reflect different planning scopes, which are known to depend on task demands and participants' abilities (Ferreira & Swets, 2002;Snyder & Logan, 2014), and not necessarily an inherent difference between language and action sequencing. ...
... The effects of repetition, however, are modulated by several other factors (Crump & Logan, 2010;Fournier et al., 2014;Jaeger et al., 2012;Logan, 2021;Snyder & Logan, 2014), and sometimes present conflicting results. Previous research on phonological onset overlap between prime and target words has shown both facilitation (e.g., Meyer & Schriefers, 1991;Smith & Wheeldon, 2004) and inhibition (e.g., Sevald & Dell, 1994;Sullivan & Riffel, 1999) effects, depending on the temporal interval between presentation of the prime and the target word (Jaeger et al., 2012). ...
... Previous research on phonological onset overlap between prime and target words has shown both facilitation (e.g., Meyer & Schriefers, 1991;Smith & Wheeldon, 2004) and inhibition (e.g., Sevald & Dell, 1994;Sullivan & Riffel, 1999) effects, depending on the temporal interval between presentation of the prime and the target word (Jaeger et al., 2012). Similar disparities have been found between typing and speaking studies (Snyder & Logan, 2014), leading to the suggestion that task demands, and in particular task speed and the producer's skill, will affect the strategy chosen for the task and the resulting outcome (Logan, 2021). This echoes our own motivation for Experiment 2, in which we found that task demands and the producer's skill can affect the scope of the plan and the resulting error pattern. ...
We investigated similarities in language and motor action plans by comparing errors in parallel speech and manual tasks. For the language domain, we adopted the “tongue twister” paradigm, while for the action domain, we developed an analogous key-pressing task, “finger fumblers.” Our results show that both language and action plans benefit from reusing segments of prior plans: when onsets were repeated between adjacent units in a sequence, the error rates decreased. Our results also suggest that this facilitation is most effective when the planning scope is limited, that is, when participants plan ahead only to the next immediate units in the sequence. Alternatively, when the planning scope covers a wider range of the sequence, we observe more interference from the global structure of the sequence that requires changing the order of repeated units. We point to several factors that might affect this balance between facilitation and interference in plan reuse, for both language and action planning. Our findings support similar domain-general planning principles guiding both language production and motor action.
... Skilled typists typed probe letters from the prime word more quickly than probe letters from another word, showing a partial repetition benefit instead of a cost. Snyder and Logan (2014) presented prime words (e.g., OCEAN) followed by a go signal (*****) or a word to type. The probe word was either identical to the prime (OCEAN), an anagram of the prime (CANOE), sharing common letters, or different from the prime, sharing no letters with it (GULPS). ...
... Thus, extensive practice did not allow the representations of the retained and intervening actions to rely entirely on LTM and so avoid code confusion. It is unclear whether more practice (such as the amount of practice in our typing example; Crump & Logan, 2010;Snyder & Logan, 2014) would have eliminated code confusion. Importantly, however, the reduction of code confusion with practice found in our study is compatible with several memory models that explain improvements in performance with practice. ...
... While we provide evidence that code confusion for offline actions can be reduced with practice, we cannot conclude that it can be eliminated with more extensive practice. The typing data suggest it can be eliminated (Crump & Logan, 2010;Snyder & Logan, 2014), but the tasks were different from the usual PRC procedures, so some uncertainty remains. However, our results make it clear that practice is an important modulator of PRCs and that motivates further research into practice effects. ...
Often, we depart from an intended course of events to react to sudden situational demands (an intervening event) before resuming the originally planned action. Executing an action to an intervening event can be delayed if the features of this action plan partly overlap with an action plan retained in working memory (WM) compared to when they completely overlap or do not overlap. This delay is referred to as a partial repetition cost (PRC). PRCs are typically attributed to code confusion between action plans in WM. We tested this by training the component action plans extensively to reduce their reliance on WM. If PRCs are caused by code confusion within WM, then PRCs should be reduced and possibly eliminated with extensive practice. To test this, participants performed a partial repetition (PR) task after 0, 4 and 8.5 sessions of stimulus-response (S-R) training. In the PR task, participants saw two visual events. They retained an action to the first event while executing a speeded action to a second (intervening) event; afterwards, they executed the retained action. The two action plans either partly overlapped or did not overlap. Results showed that extensive (S-R and PR task) practice reduced but did not eliminate PRCs. A reduction in PRCs (code confusion) with practice is compatible with memory models that assume action events become more specific and less reliant on WM with practice. These findings merit expansions of PR tasks to other domains and broader conceptions of action plans that incorporate the formal structure of memory models.
... Because reactivation is fallible, multiword repetitions are uncommon relative to single-word repetitions. We hypothesized that retrieving the past involves rewinding the chain, so that words are retrieved one-by-one, going farther and farther into the past (Harmon & Kapatsinski, 2016;Kapatsinski, 2005;Snyder & Logan, 2014). At each point, the previously reactivated words remain available to cue the past and the future, and an additional word is added to the cueing context. ...
Repetition appears to be part of error correction and action preparation in all domains that involve producing an action sequence. The present work contends that the ubiquity of repetition is due to its role in resolving a problem inherent to planning and retrieval of action sequences: the Problem of Retrieval. Repetitions occur when the production to perform next is not activated enough to be executed. Repetitions are helpful in this situation because the repeated action sequence activates the likely continuation. We model a corpus of natural speech using a recurrent network, with words as units of production. We show that repeated material makes upcoming words more predictable, especially when more than one word is repeated. Speakers are argued to produce multiword repetitions by using backward associations to reactivate recently produced words. The existence of multiword repetitions means that speakers must decide where to reinitiate execution from. We show that production restarts from words that have seldom occurred in a predictive preceding-word context and have often occurred utterance-initially. These results are explained by competition between preceding-context and top-down cues over the course of language learning. The proposed theory improves on structural accounts of repetition disfluencies, and integrates repetition disfluencies in language production with repetitions observed in other domains of skilled action. (PsycInfo Database Record (c) 2021 APA, all rights reserved).
... As we learn to type we are able to plan a complex sequence of keystrokes that reflect a hierarchical organization of motor outputs. For example, seeing a particular word can prime the constituent keystrokes, especially when they are executed in the correct order (Snyder & Logan, 2014). Eye movements in typing have been studied for many decades, with Butsch (1932) being the first to measure where people look in text while they touch type. ...
When looking at an image, participants typically concentrate their fixations on a subset of locations that can be considered salient or important. A productive avenue of research has involved trying to explain how our eyes are guided to these locations. Here, I review research that investigates fixations in more interactive tasks, where participants move around or interact with the scene rather than viewing a picture. When walking or performing actions with real objects, fixations are closely linked to the current hand and body movements. In social interactions, gaze also provides a signal that can be interpreted by other people. In sequential tasks, fixations may serve multiple functions and when and where they are guided will change with different stages of the task. Interactive tasks can be studied in real-world settings or in the laboratory, and they provide important insights into the guidance and function of gaze.
