Skip to main content

Friday Plan

  1. Introduction to Free Classification
  2. Project Summaries
  3. Stimulus Preparation
  4. Experiment Preparation
  5. Data Collection

Preparing Audio Files:

In Praat, segment your audio files to include only the speech material that you want to present to your listeners. Save each stimulus item in a separate .wav file. Once you have segmented and saved the stimulus materials, you can RMS amplitude normalize and/or downsample them in batch mode using the following Praat scripts for Mac and PC.

Making Stills:

In Powerpoint, create one different slide for each stimulus item to serve as the visual component of the stimulus movies. The participants should not be able to make interpretable groups of the auditory stimulus materials based on the visual information alone. Thus, the stills should be uniquely identifiable by the experimenter, but should not reflect the category structure of the stimulus materials themselves. Save the Powerpoint file in JPEG format. This will create a directory with one .jpg file per slide.

Making Movies:

In iTunes, import all of your audio files. In iPhoto, import all of your visual stills. In iMovie, select the Media tab to view your Audio clips in iTunes and your visual (Photo) clips in iPhoto. For each movie, select an audio clip and place it at the playhead. Select the corresponding visual image and drag it to the timeline. Drag the edges of the visual track to align temporally with the audio track. Click Play to check your movie. Export each movie using the Expert Settings to save it in .avi format for playback in Powerpoint.
To make movies on a PC, try using the freeware program VirtualDub or Windows Movie Maker. Thanks to Francisco Iacobelli for telling me about these programs.

Making Stimulus Display:

In Powerpoint, use the drawing tools to create a grid for your listeners to sort the stimulus items on. Your grid should have at least 5 times more squares than you have stimulus items to give your participants plenty of room to produce their classifications. Insert each movie into the slide and resize it to match the size of your grid. One movie should fit in each square of your grid. Arrange your stimulus materials neatly to one side of the grid. Include any relevant instructions above the grid. Your resulting stimulus display should look something like Figure 1. You can also insert your movies into the 16 x 16 grid that I have been using in my research. Present the stimulus materials to the participants using slide view, so that they can double-click to play the movies and drag them with the mouse around the screen.

Collecting Data:

Carefully consider the instructions that you give to your participants. What is the dimension that they should be using to make their classification judgments (e.g., vowel quality, regional dialect, etc.)? Can they make as many groups as they want or should they make a specific number of groups? Does each group have to have the same number of items in it? You should also provide instructions about what arrangements count as "groups" (that is, how are groups visually defined on the grid) and how many times they can listen to and rearrange the stimulus items. Sample instructions are available.