Lowit-Leuschel & Docherty (2001_

NATURE OF PROSODIC DISORDERS

ANALYSIS FORM

 

SOURCE: Lowit-Leuschel, A., & Docherty, G. J. (2001). Prosodic variation across sampling tasks in normal and dysarthric speakers. Logopedics Phoniatrics Vocology, 26, 151-164.

 

REVIEWER(S):  pmh

 

DATE: June 22, 2014

 

ASSIGNED GRADE FOR OVERALL QUALITY: B (B+ is the highest possible grade.)

 

POPULATION:dysarthria; adults

 

PURPOSE: To investigate changes in acoustic measures of prosody in the conversation and reading of speakers with dysarthria and matched peers.

 

INSIGHTS ABOUT PROSODY:

• Within group comparison of reading and conversation:

– For 3 of the outcome measures, the typical peer (TP) group performed significantly differently in conversation and reading contexts for

1. percentage of stressed vowels

2. F_o range

3. F_o variation

– The group with dysarthria (Dys) performed similarly on reading and conversation tasks for all outcomes.

• Degree of change:

– There were significant differences in the degree of change in performance in reading compared to conversation for the two groups (Dys and TP).

• Group performance characteristics and direction of change:

     – Qualitative analysis of Ps’ performance suggested the difference patterns in Dys and TP groups in conversation and reading may not be robust. It appears TD Ps were more homogeneous and that more of them changed markedly in the 2 contexts.

• Individual performance

– Qualitative analysis of individual group performance revealed no patterns of performance for Ps within either group. Performance on a single measure was not predicted from other measures.

• Comparison of Dys and TP groups:

– There was not direct statistical comparison between Dys and TP groups.

 

1. What type of evidence was identified? Prospective, Nonrandomized Group Comparison Design

 

 

2. Group membership determination:

a. If there were groups of participants were members of groups matched? Yes

b. The groups were matched for age, sex, and accent.

 

 

3. Was participants’ communication status concealed?

a. from participants? No

b. from assessment administrators? No

c. from data analyzers? No

 

                                                                    

4. Were the groups adequately described? Yes

1. How many participants were involved in the study?

• total # of participants: 32

• was group membership maintained throughout the experiment? No, 6 members of the initial Dys group and 2 members of the TP group were removed from the investigation

• # of groups: 2

• List names of groups: Dysarthria (Dys); Typical Peers (TP)

• # of participants in each group: final number (after removal of Ps noted above) in each group = 12

           

b.

– The following variables were CONTROLLED:

• age: at least 50 years of age (Dys mean = 65.9; TP mean = 66)

• gender: 6f; 6m

• accent: all from Northeast England

• cognitive skills: investigators noted this was controlled but the investigators did not provide descriptive data

• language skills: investigators noted this was controlled but the investigators did not provide descriptive data

• hearing: investigators noted this was controlled but the investigators did not provide descriptive data

• voice: Ps were excluded if voice problems affected acoustic measurement

– The following variables were DESCRIBED:

• etiology of Dys group: 6 Ps with Parkinson’s disease, 3 Ps with motor neuron disease, 3 Ps with multiple sclerosis

• severity of dysarthria: ranged from mild to severe, most Ps were mild or moderate

• intelligibility: Dys mean = 73.6%; TP mean = 97.9%

 

c. Were the communication problems adequately described? Yes.

• disorder type: (List) Dysarthria; Adult

 

 

5. What were the different conditions for this research?

a. Subject (Classification) Groups? Yes, adults with dysarthria (Dys); peers who did not exhibit dysarthria (TP)

                                                               

b. Experimental Conditions? Yes

• reading (no levels within the condition)

• conversation (no levels within the condition)

 

c. Criterion/Descriptive Conditions? No

 

 

6. Were the groups controlled acceptably? Yes

 

 

7. Were dependent measures appropriate and meaningful?

a. The dependent measures were

• Dependent Measure #1: Articulation rate (syllables per second)

• Dependent Measure #2: Mean unstressed vowel duration

• Dependent Measure #3: Number of unstressed vowels

• Dependent Measure #4: Percentage of unstressed vowels

• Dependent Measure #5: Range of intensity variation (highest and lowest peak intensity in each utterance)

• Dependent Measure #6: Intensity envelope (difference in vowel intensity for contiguous vowels in an utterance)

• Dependent Measure #7: Range of F_o variation (highest and lowest peak F_o in each utterance)

• Dependent Measure #8: F_o envelope (difference in vowel midpoint F_o for contiguous vowels in an utterance)

• Dependent Measure #9: Intra-vowel F_o variation (difference between the lowest and highest F_o for each vowel in the sample (i.e., reading or conversation)

• Dependent Measure #10: Average F_o level (overall average F_o for each the sample (i.e., reading or conversation)

 

b. None ofthe dependent measures were subjective; all measures were acoustic.

 

c. All of the dependent measures were objective as all of the measures were acoustic.

                                         

 

8. Were reliability measures provided?

a. Interobserver for analyzers? No

b. Intraobserver for analyzers? No

c. Fidelity for investigators assessment behaviors? No

 

 

9. Description of design:

• The investigators compared each group’s (Dys, TP) performance in the reading and conversation contexts for each of the 10 dependent measures.

• The investigators compared the percentage of change between the two groups (Dys, TP) for each of the dependent measures.

• The investigators qualitatively analyzed

(1) the group change data to determine if there were group (Dys, TP) differences and

(2) individual data by looking for changes in ranking among each of the dependent measures within each of the two groups.

 

10. What were the results of the inferential statistical testing?

a. The target p level was ≤ 0.01.

1. COMPARING READING VERSUS CONVERATION FOR EACH GROUP (Dys, TP)-

• Dependent Measure #3: Number of unstressed vowels: for TP group only

• Dependent Measure #7: Range of F_o variation (highest and lowest peak F_o in each utterance): for TP group only

• Dependent Measure #8: F_o envelope (difference in vowel midpoint F_o for contiguous vowels in an utterance): for TP group only

2. COMPARING Dys AND TP GROUPS FOR EACH DEPENDENT MEASURE-

The Dys group displayed significantly more change than the TP group for the following measures:

• Dependent Measure #2: Mean unstressed vowel duration

• Dependent Measure #3: Number of unstressed vowels

• Dependent Measure #4: Percentage of unstressed vowels

The TP group displayed significantly more change than the Dys group for the following measures:

• Dependent Measure #5: Range of intensity variation (highest and lowest peak intensity in each utterance)

• Dependent Measure #6: Intensity envelope (difference in vowel intensity for contiguous vowels in an utterance)

• Dependent Measure #7: Range of F_o variation (highest and lowest peak F_o in each utterance)

• Dependent Measure #8: F_o envelope (difference in vowel midpoint F_o for contiguous vowels in an utterance)

• Dependent Measure #9: Intra-vowel F_o variation (difference between the lowest and highest F_o for each vowel in the sample (i.e., reading or conversation)

• Dependent Measure #10: Average F_o level (overall average F_o for each the sample (i.e., reading or conversation)

3. The FINDING FROM 1 AND 2 APPEAR CONTRADICTORY BUT

– the investigators noted that Ps in the TP group were more homogeneous and some of the TP Ps showed greater variation in scores than did the Dys group.

 

 

b. What were the statistical tests used to determine significance?

• Mann-Whitney U:

• Wilcoxan:  

 

c. Were effect sizes provided? No

If yes, provide data and interpretation:

 

d. Were confidence interval (CI) provided? No

 

 

11. What were the results of the correlational statistical testing? There was no correlational analysis.

 

 

12. What were the results of the descriptive analysis?

• Although there were some exceptions, there were 2 major cluster performances:

(1) Ps who displayed limited change

(2) Ps who changed in a specific direction

• The differences in homogeneity between the two groups (Dys, TP) did not appear to be related to the etiology or severity of the Dys group.