Download and run the installation setup file. A password is needed to extract the included files. The password will be forwarded on request by email: "kayserj@nyspi.columbia.edu" (please state your name, institution, email address, and intended research purpose).

The setup will copy two shortcuts on the desktop: a link to the TST program and a link that will open the TST data folder. Click on the icon to run the TST program. Follow the experimenter instructions in the About box (click Help - About or <F1>).

The copyright-protected software is provided 'as is'. Its sole purpose is for non-profit scientific research. The software, whichis currently used at the Psychophysiology Laboratory of New York State Psychicatric Institute (http://psychophysiology.cmpc.columbia.edu) and at several other research institutions, is revised if bugs or problems are identified, or other modifications are needed. However, all responsibilities and consequences of using this software are completely with the user. At present, the documentation is limited to the description given below (although the interface should be largely self-explanatory). Please report bugs or any other problems to: kayserj@nyspi.columbia.edu. Downloading and installing this software implies acceptance of these terms.

Tone Properties and Test Sequence Permutation

The test requires listening to a series of tone pairs, consisting of 300-ms sine waves with frequencies between 325 and 1994 Hz separated by 100 ms silence. Half of all trials consist of identical tones (100% similarity), whereas the other half consists of different tones using five similarities (95%, 90%, 85%, 75%, 67%). The task is to determine whether the tones were the same or different. The test includes a 10-trial practice block and three 20-trial test blocks separated by 30-s rest periods. Test performance is measured by tone discrimination accuracy (percent correct) obtained for each of the tone pair frequency ratios (.67, .75, .85, .90., .95., 1.0).

Whereas the original tone screening test used a limited number of 58 different tone pairs (12 for tone similarities of 95%, 85%, 75%, or 67%, and 10 for 90%) and 20 same tone pairs, the current TST version uses a large set of 350 different (70 for each of the five frequency ratios) and 331 same tone pairs. All sound files (Waveform Audio File Format [wav], mono, 16-bit resolution, sampled at 11025 Hz) were created with Matlab R2010b scripts. Each tone was scaled to 50% maximum amplitude, with the first and last 5 ms (55 samples) of its 300-ms duration linearly tapered. To minimize the problem of "clicks" resulting from switching the sound output device in laptops (on/off), 150-ms silence periods were inserted at the beginning and end of each sound file, yielding wave files of 1000 ms duration (150 ms silence, 300 ms first tone, 100 ms silence, 300 ms second tone, 150 ms silence).

Tone frequencies were chosen using the following systematic. The (log₂-transformed) original frequency range (325 - 1994 Hz) was evenly divided into five subranges (low: 325-468 Hz; medium low: 468-674 Hz; middle: 674-971 Hz; medium high: 971-1399 Hz; high: 1399-1994 Hz). Likewise, each subrange was further divided using seven frequency steps (e.g., for the low subrange, 325, 342, 361, 380, 400, 422, and 444 Hz), with five additional subdivisions for each step (e.g., 325, 328, 332, 335, and 339 Hz). For each of the resulting 175 frequencies, difference tone pairs were constructed by computing a second tone frequency for each of the five different ratios and their inverses (e.g., 325 Hz * 0.95 = 309 Hz, and 325 Hz * 1/0.95 = 342 Hz) while excluding tone pairs resulting in frequencies outside the original frequency range (e.g., rejecting tone pair 325/309 Hz, and accepting tone pair 325/342 Hz). For each of the main five frequency subranges, an equal number of tone pairs increasing (up) or decreasing (down) in frequency were randomly selected for each tone ratio, resulting in 7 tone pairs per frequency change (up, down), frequency subrange (low, medium low, middle, medium high, high), and tone ratio (.67, .75, .85, .90., .95.). Same tone pairs were constructed from the resulting 70 difference tone pairs per frequency subrange by pairing of both the high and the low frequency tone (all frequencies were rounded to the nearest integer value), resulting in 331 same tone pairs with unique frequencies (63, 62, 68, 68, and 70 for each frequency subrange).

A pseudo-random TST sequence is created using the following constraints. Within each set of 10 consecutive trials (i.e., the practice block and each half of a test block), 5 different tone pairs, one for each tone ratio, are randomly selected from each frequency subrange, with 2 increasing (up) and 3 decreasing (down) frequencies, or vice versa (3 up, 2 down). Likewise, 5 same trials are randomly selected from each frequency subrange. Proportion of up/down trials are balanced across the two halves of each test block. Sound stimuli can only be assigned if their frequencies differ from any previously-selected frequency (i.e., any given TST sequence avoids repeating a tone frequency, other than those within a given same tone pair). Each set of 10 consecutive trials is randomly shuffled with the constraint that no more that 3 same or 3 different tone pairs occur in sequence. In this manner, each 60-trial test sequence consists of an equal number of trials for 1) same and different tone pairs, 2) each of the 5 different ratios, and 3) up and down tone pairs. In addition, an equal representation of tone frequencies across the overall frequency range for all critical conditions is maintained.

References

Bruder, G.E., Alschuler, D.M., Kroppmann, C.J., Fekri, S., Gil, R., Jarskog, L.F., Harkavy-Friedman, J.M., Goetz, R., Kayser, J., Wexler, B.E. (2011). Heterogeneity of auditory verbal working memory in schizophrenia. Journal of Abnormal Psychology, 120(1), 88-97. doi:10.1037/a0021661

Bruder, G.E., Wexler, B.E., Sage, M.M., Gil, R.B., Gorman, J.M. (2004). Verbal memory in schizophrenia: additional evidence of subtypes having different cognitive deficits. Schizophrenia Research, 68(2-3), 137-147. PMID: 15099598

Stevens, A.A., Donegan, N.H., Anderson, M., Goldman-Rakic, P.S., & Wexler, B.E. (2000). Verbal processing deficits in schizophrenia. Journal of Abnormal Psychology, 109(3), 461-471. PMID: 11016116

Wexler, B.E., Stevens, A.A., Bowers, A.A., Sernyak, M.J., Goldman-Rakic, P.S. (1998). Word and tone working memory deficits in schizophrenia. Archives of General Psychiatry, 55(12), 1093-1096. PMID: 9862552