Chapter 7: Task Performance and Individual Differences in Working Memory, Foreign
7.2 Results
7.2.1 Correlations of phonological short-term memory and complex
One-way analyses of variance (ANOVA) were conducted to reveal how the performance of students with different backward digit spans and non-word spans varies in terms of accuracy, complexity, fluency and lexical variety on the two task types in both oral and written modes. The analyses revealed
that phonological short-term memory (PSTM) assessed with the non-word span and complex working memory capacity (WMC) measured by the backward digit span affected participants’ performance differently.
As shown in Tables 43 to 46 participants’ phonological short-term memory impacted the accuracy and syntactic complexity of their output on both oral and written narrative tasks. The oral picture narration task seems to be more affected by PSTM than the parallel written task, while a similar trend may be observed for the written cartoon description task, which appears slightly more affected by PSTM than the oral version of the task. Participants with a non-word span of 9 produced significantly less error-free clauses than their peers with a non-word span of 5, 6, 7 and 8 on the oral cartoon description task. Similarly, in the case of the oral picture narration task participants whose non-word span was 9 used significantly less error-free clauses than their counterparts with a non-word span of 5, 7 and 8. In terms of the ratio of error-free relative clauses, however, participants with a non-word span of 9 produced significantly more error-free relative clauses than their peers with a NW span of 5 or 6. It is interesting to note that participants whose NW span was 6 used less correct relative clauses than the others. One task-specific syntactic complexity measure, the subordination ratio also appears to be affected by PSTM, with participants scoring higher (8-9 syllables) on the NW span test producing significantly more subordinate clauses than those who scored low (5-7).
As for the effects of PSTM on written narrative task performance, the ratio of subordinate clauses appears to be influenced by participants’ NW span scores in similar ways on both task types, i.e. those who scored 9 significantly outperformed those who repeated 5, 6, 7 or 8 syllables successfully.
Performance on the written cartoon description task shows an interesting relationship between PSTM and the ratio of error-free past tense verbs, a task-specific accuracy measure. As shown in Table 45 high NW span scorers appear to use less accurate language in terms of this measure on the written cartoon
description task. If we compare and contrast the ways in which accuracy and complexity measures are influenced by PSTM on both task types used in this study and across oral and written modalities, we may detect and interesting trend. Namely, as far as accuracy measures are concerned, almost all the significant relationships with NW span scores show that participants with a lower range score (of 5-6 syllables) produced significantly more accurate language than those with high scores (of 8-9 syllables). In contrast, students with a higher NW span produced significantly more complex narratives in terms of sentence structure than those with lower NW scores.
Table 43. The comparison of task performance of students with different nonword (NW) spans on the oral cartoon description task
Variable NW span N Mean SD F
D-value 5 2 42.07 15.16
6 3 37.83 8.56
7 21 43.78 10.29
8 15 43.73 15.62
9 3 38.21 7.39
.28
Speech rate 5 2 1.54 .18 .32
7 21 1.66 .44
8 15 1.67 .40
9 3 1.69 .20
Clause length 5 2 7.28 1.72
6 3 5.81 .52
7 21 6.24 .86
8 15 6.76 1.07
9 3 6.95 .61
1.55
Sub-ordination ratio 5 2 .21 .01
6 3 .36 .05
7 21 .36 .10
8 15 .31 .15
9 3 .37 .07
1.10
Relative clause ratio 5 2 .07 .01 1.33
6 3 .11 .10
7 21 .11 .08
8 15 .09 .07
9 3 .14 .12
Ratio of error-free Clauses
5 2 .83 .06
6 3 .79 .07
7 21 .73 .11
8 15 .80 .09
9 3 .53 .06
5.12**
5>9 6>9 7>9 8>9 7<8
Ratio of error-free relative clauses 5 2 1.02 .01 1.33
6 3 .50 .24
7 21 .62 .41
8 15 .81 .40
9 3 .33 .58
Ratio of error free past tense verbs 5 2 1.01 .01 .79
6 3 .83 .24
7 21 .75 .38
8 15 .88 .19
9 3 .64 .24
Total 44
* = p < .05
Table 44. The comparison of task performance of students with different nonword (NW) spans on the oral picture narration task
Variable NW span N Mean SD F
D-value 5 2 36.95 10.95
6 3 30.56 2.82
7 21 35.28 10.21
8 15 38.40 9.09
9 3 28.75 1.79
.96
Speech rate 5 2 1.78 .27
6 3 1.17 .19
7 21 1.65 .37
8 15 1.66 .54
9 3 1.64 .41
.95
Clause length 5 2 6.21 1.18
6 3 6.56 1.25
7 21 6.50 .64
8 15 6.92 1.05
9 3 6.38 1.66
.64
Sub-ordination ratio 5 2 .30 .06
6 3 .35 .09
7 21 .28 .11
8 15 .37 .08
9 3 .39 .10
2.43*
7<8 7<9
Relative clause ratio 5 2 .08 .02 .53
6 3 .07 .01
7 21 .10 .06
8 15 .13 .09
9 3 .14 .13
Ratio of error-free Clauses
5 2 .96 .01
6 3 .69 .13
7 21 .73 .12
8 15 .81 .13
9 3 .69 .01
2.58*
5>6 5>7 5>9
Ratio of error-free relative clauses 5 2
6 3
7 21
8 15
9 3
Ratio of error free past tense verbs 5 2
6 3
7 21
8 15
9 3
Total 44
.75 .35
.16 .29
.57 .38
.79 .21
.88 .17
1.02 .01
1.02 .01
.91 .22
.84 .34
2.91*
6<5 6<7 6<8 6<9
.39
.92 .14
* = p < .05
Table 45. The comparison of task performance of students with different non-word (NW) spans on the written cartoon description task
Variable NW span N Mean SD F
D-value 5 2 54.31 3.81 .46
7 21 65.88 21.71
8 15 63.63 14.54
9 3 54.25 4.52
Clause length 5 2 6.37 1.35
6 3 5.24 .49
7 21 5.66 .67
8 15 6.04 1.02
9 3 5.19 .41
1.39
Subordination ratio 5 2 .32 .01
6 3 .31 .08
7 21 .33 .11
8 15 .35 .07
9 3 .47 .02
3.13*
5<9 6<9 7<9 8<9
Relative clause ratio 5 2 .16 .01
6 3 .07 .11
7 21 .10 .09
8 15 .07 .04
9 3 .09 .01
.84
Ratio of error-free Clauses
5 2 .84 .07
6 3 .80 .04
7 21 .81 .09
8 15 .81 .10
9 3 .78 .07
.11
Ratio of error-free relative clauses
5 2 .83 .24 .82
6 3 .75 .35
7 21 .69 .31
8 15 .82 .32
9 3 1.01 .01
Ratio of error free past tense verbs
5 2 .97 .04 3.13*
6 3 .94 .07 5>9
7 21 .96 .07 6>9
8 15 .96 .06 7>9
9 3 .65 .56 8>9
Total 44
* = p < .05
Table 46. The comparison of task performance of students with different nonword (NW) spans on the written picture narration task
Variable NW span N Mean SD F
D-value 5 2 67.01 8.69 .47
6 3 57.07 2.27
7 21 55.76 15.94
8 15 62.41 19.91
9 3 57.06 7.42
Clause length 5 2 6.22 .59
6 3 6.01 1.40
7 21 6.27 1.02
8 15 6.20 .79
9 3 6.36 1.02
.06
Sub-ordination ratio 5 2 .31 .13
6 3 .27 .04
7 21 .32 .11
8 15 .41 .12
9 3 .41 .15
1.76*
7<8
Relative clause ratio 5 2 .07 .03 .27
6 3 .13 .16
7 21 .14 .08
8 15 .14 .09
9 3 .11 .04
Ratio of error-free
Clauses 5 2 .69 .13
6 3 .86 .08
7 21 .83 .11
8 15 .82 .08
9 3 .75 .09
1.41
Ratio of error-free relative
clauses 5 2 .50 .70 1.22
6 3 .94 .09
7 21 .64 .34
8 15 .75 .22
9 3 .52 .04
Ratio of error free past tense verbs
5 2 .96 .05 1.10
6 3 .95 .06
7 21 .96 .05
8 15 .98 .03
9 3 .93 .09
Total 44
* = p < .05
Tables 47 to 50 display the effects of complex working memory capacity (WMC) as assessed by the backward digit span on oral and written narrative task performance. As shown, the written tasks seem
to have been slightly more affected by WMC than the parallel oral tasks. Working memory capacity affected the syntactic complexity of the participants’ output in the oral cartoon description task and both written task types. In addition to structural complexity, WMC has also affected lexical diversity in the written cartoon description task.
In the oral cartoon description task, students with a backward digit span of 5 produced significantly shorter clauses than learners with a backward digit span 7. In terms of the ratio of sub-ordinate clauses per AS-unit, however, students with a backward digit span of 5 used significantly more sub-ordinate clauses than their peers with a backward digit span of 4 and 7. In the cognitively more complex oral picture narration task, however, working memory capacity did not seem to play any role. In writing BW span scores affected task-specific syntactic complexity measures in both task types.
Participants with higher BW span scores (6-7) appear to have used significantly more subordinate clauses than their peers with lower scores of 4 and 5 in the written cartoon description task. Lexical variety on the same task seems to be impacted by WMC in a similar way, that is, those who scored 7 on the BW span test used significantly more varied vocabulary than the other participants. In the case of the written picture narration task, however, participants with a score of 5 produced significantly more relative clauses than their peers with a backward digit span of 4, 6 and 7.
In terms of both measures of working memory: non-word span, the measure of phonological short-term memory and the backward digit span assessing complex WM capacity, it can be observed that task-specific measures of linguistic performance seemed to reveal more statistically significant effects than widely used general measures.
Table 47. The comparison of task performance of students with different backward digit spans on the oral cartoon description task
Variable Digit span N Mean SD F
D-value 4 13 40.12 10.22
5 9 42.70 5.89
6 12 47.75 15.63
7 10 40.46 11.16
1.26
Speech rate 4 13 1.58 .23
5 9 1.53 .48
6 12 1.62 .28
7 10 1.79 .59
.71
Clause length 4 13 6.38 1.02
5 9 5.82 .54
6 12 6.52 .89
7 10 7.09 1.10
2.87*
7>5
Sub-ordination ratio 4 13 .28 .12 3.42*
5 9 .41 .08 4<5
6 12 .37 .08 7<5
7 10 .28 .13
Relative clause ratio 4 13 .09 .08 .21
5 9 .12 .08
6 12 .11 .07
7 10 .09 .08
Ratio of error-free Clauses
4 13 .74 .12 .11
5 9 .76 .11
6 12 .73 .12
7 10 .75 .11
Ratio of error-free relative clauses
4 13
5 9
6 12
7 10
Ratio of error free past tense verbs
4 13
5 9
6 12
7 10
Total 44
.56 .45
.68 .38
.81 .33
.57 .53
.76 .38
.87 .14
.88 .19
.76 .36
.77
.48
* = p < .05
Table 48. The comparison of task performance of students with different backward digit spans on the oral picture narration task
Variable Digit span N Mean SD F
D-value 4 13 38.41 8.34
5 9 29.91 9.92
6 12 36.36 10.07
7 10 38.17 8.07
1.81
Speech rate 4 13 1.67 .31
5 9 1.47 .51
6 12 1.56 .40
7 10 1.67 .50
.53
Clause length 4 13 6.83 .95
5 9 6.26 .69
6 12 6.41 .82
7 10 6.63 .75
1.04
Sub-ordination ratio 4 13 .29 .11 .79
5 9 .35 .11
6 12 .34 .09
7 10 .31 .10
Relative clause ratio 4 13 .12 .07 1.10
5 9 .08 .05
6 12 .13 .09
7 10 .10 .06
Ratio of error-free Clauses
4 13 .71 .16 .85
5 9 .76 .10
6 12 .82 .08
7 10 .74 .12
Ratio of error-free relative clauses
4 13 .75 .31 1.16
5 9 .46 .46
6 12 .56 .34
7 10 .66 .37
Ratio of error free past tense verbs
4 13 .89 .29 .73
5 9 .98 .02
6 12 .90 .28
7 10 .80 .30
* = p < .05
Table 49. The comparison of task performance of students with different backward digit spans on the written cartoon description task
Variable Digit span N Mean SD F
D-value 4 13 60.48 16.44
5 9 58.49 15.61
6 12 58.74 17.75
7 10 75.61 15.38
2.54*
4<7 5<7 6<7
Clause length 4 13 5.72 .93
5 9 5.94 1.14
6 12 5.69 .74
7 10 5.75 .64
.16
Sub-ordination ratio 4 13 .31 .06
5 9 .31 .11
6 12 .36 .11
7 10 .40 .07
2.51*
4<7 5<7
Relative clause ratio 4 13 .08 .05 .37
5 9 .09 .12
6 12 .11 .08
7 10 .07 .04
Ratio of error-free Clauses
4 13 .80 .09 .38
5 9 .81 .09
6 12 .79 .07
7 10 .83 .09
Ratio of error-free relative clauses
4 13 .81 .32 .65
5 9 .64 .39
6 12 .84 .23
7 10 .75 .28
Ratio of error free past tense verbs
4 13 .96 .06 1.07
5 9 .95 .06
6 12 .86 .30
7 10 .98 .03
Total 44
* = p < .05
Table 50. The comparison of task performance of students with different backward digit spans on the written picture narration task
Variable Digit span N Mean SD F
D-value 4 13 57.34 12.89 .46
5 9 54.45 15.64
6 12 60.67 19.10
7 10 62.64 18.34
Clause length 4 13 6.14 .64 .79
5 9 6.35 .92
6 12 5.96 .88
7 10 6.54 1.21
Sub-ordination ratio 4 13 .33 .11 .21
5 9 .35 .08
6 12 .37 .16
7 10 .36 .12
Relative clause ratio 4 13 .09 .06 3.39*
5 9 .21 .10 5>4
6 12 .12 .07 5>6
7 10 .13 .08 5>7
Ratio of error-free Clauses
4 13 .82 .08 .12
5 9 .81 .13
6 12 .83 .11
7 10 .82 .09
Ratio of error-free relative clauses
4 13 .66 .38 .16
5 9 .74 .25
6 12 .65 .31
7 10 .71 .28
Ratio of error free past tense verbs
4 13 .98 .06 .63
5 9 .96 .10
6 12 .95 .07
7 10 .96 .08
Total 44
* = p < .05