Article Summaries & Critiques
HOME	LITERATURE REVIEW	REFLECTIONS & HYPOTHESIS	TEMPORAL PROCESSING	STUDY DESIGN	REFERENCES	ARTICLE SUMMARIES	HOTLINKS

	Ellis, Mark C. (1992) Tempo perceptions and performance of elementary students Grades 3-6
	Swanson, H.L, and Berninger, V.W (1996) Individual differences in children’s working memory and writing skill
	Willingham, D.B. (1998) A neuropsychological theory of motor skill learning
	Yates, Cheryl M, Berninger, V.W., Abbott, Robert D. (1995) Specific Writing Difficulties in intellectually gifted children (Critique)

Ellis, Mark C. (1992)Tempo perceptions and performance of elementary students Grades 3-6  Journal of Research in Music Education 40(4) pp.329 341.

 
 

Research questions:

a) Do tempo perception and performance abilities increase in a linear fashion from Grade 3 to Grade 6?
for conclusion

b) Do subjects with more training perceive and perform tempo better than those with less training?
for conclusion

c) Does the tempo at which tasks are presented affect tempo perception and performance?
for conclusion

d) To what extent are tempo perception and performance abilities related?
for conclusion
 

Method:

Participants:
80 subjects, 20 randomly selected from each of grades 4-6, at a midwestern elementary school, in a predominantly middle class area, but which had a minority of students bussed in from outside the area.  All students receive 30 minutes general music instruction three times per week; 80% of students begin instrumental instruction by grade 4, and 50% of students in grades 3-6 have received piano instruction.

Procedures:
Tempo perception:
1) tempo discrimination: subjects were asked to detect gradual changes in the tempo of a short melody.  Three initial tempi were used, and examples either increased or decreased in tempo, by varying amounts.  Subjects had to indicate on an answer sheet whether the melody had increased, decreased, or stayed at the same tempo.  There were 36 presentations.

2) subjects listened to 30 pairs of a rhythm pattern, and indicated on an answer sheet whether the item was the same, faster, or slower than the first.

Tempo performance:
Subjects had to continue a rhythmic pattern by tapping on a synthesizer key.  Subjects were asked to begin “in tempo”, “without delay”, and to play “as steadily as possible”.

Accuracy of tempo was determined by calculating the time from first to last keystroke, then comparing that time to the time that would have elapsed had the tempo been exact.

Steadiness of tempo was determined by calculating the average deviation of each keystroke from the time it should have occurred, in his or her own tempo.
Response “in tempo” was assessed thus: a performance was considered to have been “in tempo” if the initial keystroke occurred within + or – 25% of where it should have occurred.

Subjects were allowed to familiarize themselves with the task before testing. MIDI equipment was used for both stimuli and recording of performance.

Data analyses:
Subjects from each grade level were assigned one point for each year of private piano lessons and each year of instrumental lessons.  Subjects at each grade level were then divided into high and low training groups by median splits.  ANOVA was performed for the two perception and two performance measures.

Results

a) Do tempo perception and performance abilities increase in a linear fashion from Grade 3 to Grade 6?
Patterns of scores on the perception tests showed a linear increase with grade, rising from scores for grade 3 on the tempo change perception test that were scarcely better than chance.
Steadiness increased linearly with grade from 3-5 and were stable from 5-6.
“In tempo” responses increased from 0% in Grade 3; 25% for Grade 4; 75% for Grade 5 and 95% for Grade 6, showing a marked linear increase.
No linear pattern was found for tempo accuracy.
return to question

b) Do subjects with more training perceive and perform tempo better than those with less training?
Training was significant or Grade 3 scores on change perception and for Grade 6 on accuracy. However, the differential contributions of maturity and training on the correlation between perception and performance were unclear because of confounding.
return to question

c) Does the tempo at which tasks are presented affect tempo perception and performance?
Performance accuracy increased as tempo increased; subjects more often performed to fast than too slow.  The ratio of faster to slower performances decreased as tempo increased.
return to question

d) To what extent are tempo perception and performance abilities related?
Correlation between perception and performance was moderately strong, and increased with grade level. However, the differential contributions of maturity and training were unclear because of confounding.
return to question
 

 
 

Swanson, H.L, and Berninger, V.W (1996) Individual differences in children’s working memory and writing skill Journal of Experimental Child Psychology 63 pp.358-385.

 
 
 

Research questions:

a) Are individual differences in working memory and writing related to a general or a process-specific system?
Background to question: Theory states that working memory (WM) has processing and storage functions that compete for capacity (Just and Carpenter, 1992).  When WM capacity limits ability in reading or writing (Beal, 1990), is it because of general storage capacity limitations (Turner and Engle, 1989) or task-specific processing efficiency limitations (Daneman and Carpenter, 1980)?
for conclusion

b) Do working memory tasks operate independently of phonological short term memory on measures of writing and reading?
Background to question: Working memory (WM) is measured using tasks that require the holding of a small amount of information while carrying out further operations.  Short term memory (STM) is measured using tasks where the information is held passively and reproduced untransformed. Differences between WM and STM that might suggest non-overlap: WM is related to high-order cognition, STM less so (Daneman and Carpenter, 1980); transcription process may use lower-order skills than text generation processes (Hidi and Hildyard, 1983); STM reflects the use of a phonological (articulatory) code, WM does not (Salame and Baddeley, 1982).  Flower and Hayes (1980) imply that “the skilled writer must have flexible access to a wide range of mental representations of actual and intended text” suggesting interaction between STM and WM, even though they may nonetheless be separate.
 

WM	STM
capacity to hold information while carrying out further operations	capacity to hold information passively and reproduce it untransformed
related to high-order cognition	less related to high-order cognition
does not use phonological (articulatory) code	uses phonological (articulatory) code

for conclusion

c) Does working memory predict variance in writing beyond that predicted by reading alone?
Background to question:  No previous studies make compare the relationship between memory and writing to that between memory and reading. Reading involves input, writing primarily output – STM may therefore make a bigger contribution to reading.
for conclusion
 

Operational definitions:

Working memory tasks: those that require some processing, inference or transformation as well as storage.
Short-term memory tasks: those that require the storage of information with minimal “ongoing”processing  requirements.

Method:

Participants: 100 fourth graders, 100 fifth graders and 100 sixth graders from five urban schools in greater Seattle. 50 girls and 50 boys were included at each grade.  Sample was ethnically and socio-economically diverse, and mean Verbal IQ score (WISC-R) was 108.4 (SD + 12.93)

Procedures:

The test battery consisted of measures of:
 

Verbal working memory:
Listening recall sentence span
Listening generation sentence span
Rhyming
Semantic categorization
Semantic association
Phrase sequence
Story recall

Visual Spatial working memory:
Visual matrix
mapping and directions

Short term memory
phonetic memory
written recall of non words
written recall of letter clusters in nonwords

Measures of composition
narrative frame
expository frame

handwriting
transcription
spelling

Vocabulary
Vocabulary Subtest of WISC-R (age-corrected)

Reading (from Woodcock Reading Master Test – Revised, age corrected)
Word Identification (real words)
Word Attack (reading nonwords)
Reading comprehension)

Each child participated in three sessions, consisting of two group sessions (approximately 20 children) 45 minutes in duration, and one individual session, 45 minutes in duration.

Session 1 (group): writing measures
Session 2 (group): listening recall, listening generation, visual matrix, map directions
Session 3 (group): remaining tasks.

Data analyses:
Means and standard deviations for all independent and dependent measures were calculated, and for all subsequent analyses, all scores were converted to z scores.  Memory measured were factor analyzed with a principal component analysis, via a varimax rotation, using Kaiser’s criterion and Cattell’s scree test.
 

Results

Four factors were found:

Factor 1:  interpreted as reflecting phonological STM.  All phonological STM task loaded on this factor.

Factor 2:  may reflect contribution of verbal WM processing Verbal WM tasks loaded on this factor.

Factor 3: may reflect contribution of executive processing in the verbal domain. Sentence span, semantic association and categorization measures loaded on this factor

Factor 4:  may reflect contribution of visual-spatial WM processing.  Visual span measures and temporal sequencing measures.

Additional findings:
Memory measures:
· older children performed better  than younger children on memory measures.
Writing measures:
· older children were better than younger children on writing measures;
· females performed better than males.

Conclusions:

a) Are individual differences in working memory and writing related to a general or a process-specific system?

Writing is distinguished from reading by its reliance on a process-specific capacity of WM, involving executive processing within a verbal system.  In contrast, a general memory system “does capture reading comprehension performance”.
return to question

b) Do working memory tasks operate independently of phonological short term memory on measures of writing and reading?

“Factor analysis indicated that WM and STM tasks did not share a similar factor.”  WM predicted text-generation in writing better that did STM, while STM predicted text transcription better than did WM.
return to question

c) Does working memory predict variance in writing beyond that predicted by reading alone?

“Working memory was found to be a significant predictor of writing when entered into the regression equation, even when reading was entered into the equation first.”return to question

 
 

Willingham, D.B. (1998) A neuropsychological theory of motor skill learning Psychological Review 105(3) pp.558-584.

This article outlines a neuropsychological theory of motor skill learning.  It is not an empirical study.

Definitions:

Motor control: the planning and execution of movement
Motor skill learning: the increasing spatial and temporal accuracy of movements with practice.

Three hypotheses:

1.  That motor skill learning grows directly out of motor control processes (giving the theory its name COBALT –  control-based learning theory).
2. That there are four processes that support motor control, which are neurally and anatomically distinct, and which utilize different forms of cognitive representation: allocentric space; egocentric space; and muscle innervation:

* COBALT currently does  not account for timing accuracy, only for spatial accuracy.  Words in italics are my additions.

3. That two mechanisms may operate during motor skill learning, corresponding to either a conscious or an unconscious mode of execution:
 

• Unconscious: as the task is practised, one or more of the four motor control processes becomes more attuned to a particular task, resulting in learning.

• Conscious: that the conscious strategic process may substitute for some of the other processes during motor skill learning, leading to improved performance.

Willingham speculates that while skills are being learned, the use of the conscious mode results in improved performance, while simultaneously allowing the unconscious processes to become attuned to the task.  However, use of the conscious mode once skills are learned may result in the phenomenon of “choking”.
He postulates that the following consequences flow from the use of a conscious or unconscious mode of execution:

Limitations of theory:

As Willingham does not claim that COBALT addresses temporal aspects of motor skill learning, he does not address learning of skills that involve timing, or moving targets.  He points out that the only inclusion of temporal processing in the theory is in the sequencing process, in which time is represented on an ordinal scale.  He writes: “…more precise temporal information on a ratio scale seems likely to be crucial for many motor skills, particularly those where the actor does not control the timing of the task.”  Integrating such temporal information into the existing model is an important challenge for future research.

Yates, Cheryl M, Berninger, V.W., Abbott, Robert D. (1995) Specific Writing Difficulties in intellectually gifted children Journal for the Education of the Gifted 18 (2) pp 131-155

This study sought to discover whether verbally gifted children who show excellence in writing excel in both text generation and text transcription skills, or in only one of the two.

The authors attribute lack of previous study in this area to a number of factors, including the difficulty people have in accepting that gifted children can have learning disabilities; that gifted children may not be gifted in all domains; and the fact that writing is a complex and heterogeneous domain.

Design
A total of 120 children were studied, consisting of 10 gifted and 10 average children from each of Grades 1 to 6.  Gifted children were defined as achieving Verbal IQ scores on the WISC-R of 122 or greater.  The average children were matched by age and gender to the gifted group, and selected so that they formed a homogeneous group with Verbal IQ scores of as near to 100 as possible.
A writing test battery was designed to differentiate between ability to generate text (to put ideas into words) and to transcribe text (to translate those words into written text).  The tests assessed six component skills of writing: handwriting, spelling, compositional fluency, sentence complexity, vocabulary sophistication and compositional quality.  Developmental skills related to writing, such as neuro-motor finger function and orthographic coding skills were also assessed.

Results
The authors found that on measures of text transcription ability, there was no significant difference between the gifted children and their non-gifted peers.  These measures included measures such as word count,  number of clauses, and sentence complexity, which the authors term measures of “low-level text transcription ability”.  However, on measures of text generation ability, which included vocabulary sophistication, and compositional quality ratings, the gifted group were significantly better than their average peers.
Results were analysed to ascertain the incidence for each group of writing disability in at least one area, using both relative and absolute measures.  Overall, on relative criteria (relative to verbal IQ), 43.3% of the gifted group and 8.3% of the average group were found to have a writing disability. On absolute criteria (5% percentile), 11.7% of gifted group were found to have a disability, and 25% of the average group.

Conclusion
The authors conclude that the study supported their hypothesis that the differentiation between text generation skills and text transcription skills is a valid one, and that specific writing disabilities in verbally gifted children are more likely to reflect deficits in low- rather than high-level skills. They draw three inferences for teachers of gifted students: that gifted students’ text generation potential may be masked by merely average, or below average transcription skills; that such children may suffer high frustration levels as a result of the mismatch, resulting in loss of motivation and self-esteem, task avoidance, or increased verbal precocity; and that explicit remediation may be required to deal with these disabilities.  They point out that in general, programs for gifted writers focus on high level thinking and problem solving skills, and not on low-level mechanical skills, whereas it may be help with the latter that enables the former to be expressed.

Value and validity
Three valuable implications flow from the study.  The first concerns the finding of a substantial incidence of relative writing disabilities in gifted children. While there may be strong arguments for using absolute criteria for diagnosing a specific learning difficult (Seigel, 1989, Stanovich, 1989), the fact that writing is likely to be a highly important expressive medium for gifted children means that even a relative writing disability in a verbally gifted child presents an important educational problem.  This study shows the extent of its prevalence. Though the numbers in each grade sample were small, and the percentages and means for each grade must therefore be treated with caution, the overall differentials reported were sufficiently large that the the finding of a substantial proportion of relative writing disabilities in a group verbally gifted children would seem to be valid.

The second implication concerns the fine grain of the problem. The fact that a differential emerged between low-level text transcription ability and high-level text generation ability suggests the development of detailed diagnostic and remediation programs that match the intervention to the level (text transcription or text generation) of any writing disability. As regards the validity of the finding, the fact that the test battery was carefully designed to tap a great many aspects of the writing ability construct suggests that the findings that text transcription and text generation are separable aspects of the writing process is robust.

The third implication concerns what is perhaps the most intriguing extrapolation one might make from the findings: that when writing ability is dissected into text generation and text transcription components, the former is revealed as an indicator of verbal IQ, the latter as an indicator of maturation.  This would have important implication for the handling of writing instruction in young gifted children, particularly those showing strongly asynchronous development.  Although the finding that measures of high-level text generation varied with ability rather than with grade is perhaps unsurprising in a sample with 30+ verbal IQ points separating the two groups, what is more striking is the finding that, despite such a major discrepancy between the verbal IQ levels of the two groups, measures of low-level text transcription nonetheless tended only to vary with grade level, and not with ability level.

A weakness of the study is its restriction to verbally gifted children and an artificially homogeneous “average” group. This means that only limited generalisation can be made to other more heterogeneous groups. However, the more heterogeneous control group that would have resulted from a random sample would have been more likely to mask the emergence of what may be a developmental aspect of writing skill that is independent of giftedness in text generation, which is possibly the most important finding of the study.