lab代写 – 这是lab代写的题目
AX Discrimination In lab 2, you discovered that English-speaking participants identified a [ba da ga] continuum in a categorical manner: little within category variation of identification responses and a sharp identification drop-off at category boundaries. For example, at ba 6 , there was an abrupt de- crease in b responses and a steep increase in d responses. This good within a category identification and poor between category identification is a hallmark of categorical perception. In the ABX discrimination experiment, you also discovered that participants are better at discrimi- nating pairs of sounds on different sides of a category boundary (e.g., ba 5 -ba 7 ) compared to those that fall on the same side (e.g., ba 1 -ba 3 ). Again, this good between category discrimination and poor within category discrimination is another hallmark of categorical perception. For Lab 3, you are going to extend this research to a series of different contrasts. You performed four short AX discrimination tasks. In each experiment, you heard a pair of sounds and your task was to determine if the second sound was the same or different from the first. The inter-stimulus interval was 500 ms and inter-trial interval was 1000 ms. Each pair was counter- balanced in their order of presentation and a total of 72 trials were presented during each exper- iment32 same trials and 32 different trials.
- Experiment 1 (allstopsone): English voiceless unaspirated [p]^1 and aspirated [p] pair.
- Experiment 2 (allclickstwo): lateral click [] and alveolar click [].
- Experiment 3 (allstopstwo): Spanish/Russian pre-voiced [b] and voiceless unaspirated [p].
- Experiment 4 (allclicksone): palatal-alveolar click [] and alveolar click [].
Download the four result files from Blackboard: Labs > Lab_3 > Lab 3 Data Files.
Assignment
- Provide one bar plot comparing the same and different responses (in different coloured bars) for each experiment. You should plot proportion correct and standard error bars. Label your axes.
- Provide the means and standard deviations of the proportion correct for the same and dif- ferent responses for each experiment in a table. Arrange your table as such:
- Repeat steps 1 and 2 for Reaction Times. Provide a plot and table as above.
- Reason about the inverse correlation between accuracy and RT. Is this relationship is ex- pected? Do we expect high accuracy to correlate with lower mean lower reaction times. Why or why not? Compare the stop experiments independently from the click experiments. (~ 1 paragraph)
Experiment Different (mean) Different (sd) Same (mean) Same (sd)
[p] - [p]
[b] - [p]
[] - []
[] - []
(^1) We usually use [b] to refer this phonetic segment in English, but because we are reserving [b] for the pre-voiced bilabial stop, we use [p] here, which is the more technical symbol for the English unaspirated/voiced stop.
- Compute the d scores for each experiment, using the formula: d = z(Hits) – z(False Alarms). Present these in one bar plot where d is plotted by experiment (that is, there should be one plot with four bars). This should be done on an individual-by-individual basis first before tak- ing the overall d.^2
- How do the d prime scores match up with your table in Q2?
- Now, create a plot that provides the d scores for each experiment but also for each partici- pant this should be one large plot with with 15 smaller bar plots, one for each participant. How many participants follow the qualitative pattern you noticed in Q5 and how many do not.
- Finally, consider why participants did so well in Experiment 1 and relatively well in Experi- ments 2 and 4 compared to Experiment 3. What might you hypothesize about second lan- guage sound perception based on these results? Consider whether the categories exist in the first language or not. How might your first language impact perception of native and non- native categories?
(^2) This does not mean you need to do this separately for all participants, but can be done using ddply() in two steps, first by includ- ing the participants in the formula, then with that data, excluding subject from the formula. Consult the in-class R notes.