Exam Feedback

Question 0

How agreement of perceptual ratings is assessed?

Answer 0

• intra observer and inter observer reliability: comparison between 2 raters of a set of voices/ standard bivariate Pearson correlation, quantifies the degree to which observers are rating in a parallel fashion with similar standing among cases
• intra and inter observer agreement: agreement evaluates the extent to which identical scores are assigned to each voice by each of the raters/ high level implies same meaning for each rater
• inter reliability and agreement more than 2 raters: scale score is determined reliability is calculate, to evaluate reliability calculate the mean of pair wise correlation/ confidence intervals or SE can indicate consistency or agreement among raters

Question 1

Four methods of perceptual scaling

Answer 1

1. Rating scale with equal appearing intervals
2. Visual analogue scale- end points are defined
3. Direct magnitude estimation: provided with a voice that acts as an anchor
4. Paired comparisons: statistical techniques are used to determine the perceptual distance of cases from each other/ mostly used in research

Question 2

Coefficient of variation

Answer 2

Measure of variability- tells us how large the SD is relative to the mean
Lower variability the better

Question 3

Evaluation of how well sample means represent population means

Answer 3

Lower SEM better estimate of population: SEM is inversely proportional to the sample size

Question 4

Strategies to improve sampling

Answer 4

Increase sample smaller the sample larger the sample error

Although a larger sample more time and effort needed to collect trade effort for accuracy