Double Staircase Method Research Articles

Assessment of interlinked double staircase acuity test

PurposeTo compare a novel visual acuity threshold algorithm using a predictive double staircase method with standard ETDRS logMAR chart.MethodsSoftware was written in MATLAB to generate a computerised visual acuity test. The algorithm uses a randomly presented double staircase method, the staircases are interlinked using a prediction of threshold based on each prior response. This reduces test duration and increases accuracy. The screen was calibrated using a Vernier scale and there solution of the test was 0.02 log units. Ten threshold measurements were made for each test, the results expressed as the mean and standard deviation. The test was performed by three observers each with an induced range of refractive errors and compared with the result ofETDRS logMAR acuity using a chart generated with the same screen calibration.ResultsMean stair cased acuities were −0.13, 0.02, 0.22, 0.36, 0.56 log unitsfor 0, 0.5, 1, 1.5 and 2 dioptres of defocus respectively. The mean single letter score logMAR acuities were −0.08, 0.01, 0.17, 0.38, 0.58 log units. The mean error in the stair cased test was 0.02 (SD 0.01) log units forall defocus values. The mean difference between the two tests was −0.01 log units (SD 0.09). Bland‐Altmann plots showed no systematic difference. The mean time to perform the staircased test was 93 seconds (SD 19s).ConclusionsThe interlinked double staircase method gives a fast and accurate assessment of visual acuity which compares well with ETDRS logMAR.Ten threshold measures allow error estimation which is not available in a logMAR test. Other advantages over logMAR include a true resolution of 0.02 log units rather than interpolation and random letter generation with no memory effect.

Effects of similarity on apparent motion and perceptual grouping.

Effects of similarity in colour, luminance, size, and shape on apparent motion and perceptual grouping were examined in part 1 in two parallel experiments on the same seven subjects. In both experiments, the effect of similarity was compared with that of proximity in competitive, bistable stimulus situations. A combination of a larger horizontal separation between the homogeneous stimulus elements and a smaller constant vertical separation between heterogeneous stimulus elements produced two kinds of apparent motion (or perceptual grouping) with equal probabilities. Such matched separations between homogeneous stimulus elements were obtained by the double staircase method in various stimulus conditions. In both experiments on apparent motion and perceptual grouping matched separation was found to increase as the difference between the heterogeneous stimulus elements increased. High correlations (0.71 to 0.94) of matched separations were found between apparent motion and perceptual grouping in four stimulus series: colour, luminance, size, and shape. Six of the seven subjects were also tested in part 2. Here, the effects of differences were found to work additively across different perceptual attributes in both phenomena, when multiple differences were combined in heterogeneous elements. The experimental results are discussed from the point of view that apparent motion is an example of perceptual constancy.

Temporal-discontinuity detection with contrast-modulated gratings.

Three experiments on temporal-discontinuity detection were carried out. In experiment 1, temporal-discontinuity thresholds were measured for sinusoidal gratings by the use of the double-staircase method. A sinusoidal grating was presented twice successively. The subject judged whether or not an interval was present. The temporal-discontinuity threshold increased as the spatial frequency of the grating increased, but decreased as the contrast of the grating increased. In experiment 2, contrast-modulated gratings were used instead of the sinusoidal grating. The temporal-discontinuity threshold increased as the carrier frequency increased, and the threshold for each contrast-modulated grating was similar to that for the no-modulation (sinusoidal) grating whose contrast was the same as the maximum local contrast of the contrast-modulated grating. In experiment 3, temporal-discontinuity thresholds were measured for low-contrast (3%) sinusoidal gratings. The thresholds were very low, even for such low-contrast gratings. These results suggest that the low-spatial-frequency channels are not involved in detecting the modulation frequency of the contrast-modulated grating. Rather, the local contrast seems to be the determinant of the detection of the contrast-modulated grating itself.

THE INFLUENCE OF METHOD ON THE STABILITY OF DARK FOCUS POSITION OF ACCOMMODATION

Abstract The stability of the dark focus position of accommodation was investigated using a Badal laser optometer. A quick method (double staircase method) and a longer method (method of constant stimuli) were employed. Diurnal, daily and test – rerest data were obtained. A method of limits study was included on a limited number of observers. Method of constant stimuli and method of limits gave small but significantly higher estimates of DFP. The results confirm the conclusions of other workers that the dark focus position of accommodation is a highly stable ‘trait’‐like property of the human visual system. The most efficacious and valid method for measuring dark focus position of accommodation was found to be the double staircase method.

Comparison of Loudness Balances and Coupler Measurements for Insert-Type and Supra-Aural Earphones

Loudness balances were performed between a supra-aural earphone and an insert-type earphone. Bandlimited white noise was used as the signal. Relative balances were obtained, with both test earphones sequentially balanced against a standard earphone. A double staircase method with variable step size was used to obtain the balances. The insert-type earphone used a plastic tube with ear tip to couple to the ear canal. Each of the ten subjects balanced with the properly fitted tip, and with one tip size larger and smaller. Objective loudness measurements were performed using the Electro-Acoustic Rating System (EARS) [Sullivan, Bell Syst. Tech. J. 50, 2663–2741 (1971)]. A modified type 3A coupler was used [Bryant, J. Acoust. Soc. Am. 52, 1599–1606 (1972)]. Since ear-insert couplers measure the sound pressure at the eardrum, and since the standard supra-aural coupler measures the pressure at the ear canal, a special “tympanum” coupler (Bryant, 1972) was used for the supra-aural earphones. The results of the objective measurements showed the ear-insert type to be 2.6 dB louder than the supra-aural type, which was in good agreement with the 2.5 dB obtained by the loudness balances, and supports the use of the “tympanum” coupler for comparative measurements.

Two-flash threshold in normal and subclassified schizophrenic groups.

A normal and various subclassified schizophrenic groups ( ns = 10) were compared on two-flash threshold by a brief double staircase method and by an extensive yes-no procedure with signal-detection analytic techniques. For all groups the two techniques correlated moderately. Contrary to predictions made by Venables' arousal theory, the data obtained from the yes-no procedure showed that these schizophrenic groups performed similarly and at a lower sensitivity than the normal group.

Speech Quality of Computer-Simulated Voice-Switched Amplifiers

Voice-switched amplification can be used in telephone circuits to provide additional amplification in the direction of transmission only when speech is present and thus increase the apparent S/N ratio. The effect on speech quality is a function of operating characteristics such as attack time, release time, switching thresholds, and amount of switched amplification. To evaluate the detectability and speech quality with such amplifiers, subjective tests were conducted in which computer simulation was used to provide convenient and accurate control of the operating characteristics. In the detection tests, the subjects determined whether they could hear the operation of the voice-switched amplifier, using a modified double staircase method with variable step size. The results were used to map out a region in which the operation of the switched-gain device was not detectable. A category judgment procedure using a 5-point scale was used for the transmission quality listening tests in the previously determined detectability region. Decay times of 25, 100, and 300 msec were presented with switched gains of 0, 10, 15, and 20 dB, over circuits with S/N ratios of 10 and 30 dB. Results indicated that for a 10-dB S/N ratio, the effects of amount of gain switched and decay times were both significant and had a strong interaction. For a 30-dB S/N ratio, only the effect of decay times was significant. The 10-, 15-, or 20-dB switched gain conditions were always preferred over the 0-dB condition.

Loudness Tracking and the Staircase Method in the Measurement of Adaptation

The results of several auditory adaptation experiments suggest that binaural interaction, resulting from comparison stimuli presented to the control ear, may affect the amount and rate of adaptation. An attempt was made to evaluate the effects of such interaction with a tracking procedure. The tracking method was found to have 3 shortcomings. Firstly, subjects could predict the direction of attenuation of the comparison signal. Secondly, they could produce an apparently acceptable tracing by responding in a rhythmic fashion or by counting the comparison signals. A third problem appeared to exist in that the tracings moved in the direction of negative adaptation at the start of the adapting period. The double staircase method described by Cornsweet [1962] was employed in an attempt to overcome the disadvantages of tracking. The staircase method was successful in preventing the subjects from predicting the loudness of succeeding comparison stimuli. It also discouraged counting and rhythmic responses but did...

Loudness Adaptation: Use of the Double Staircase Method and Heterophonic Balances

Twelve normal-hearing listeners were adapted monaurally with a 4-kHz pure tone of 60 dB SPL for 5 min following 3 min of preadapting loudness balances. The control ear was simultaneously stimulated every 3 sec with a 0.25-, 1-, or 4-kHz comparison tone of 1- or 2-sec duration. The listener was instructed to compare the loudness of each comparison signal with that of the adapting signal. Comparison signal intensities were controlled by the listener through two recording attenuators with the comparison signal randomly switched to one or the other attenuator. The magnitude of adaptation was calculated at 20-sec intervals. Average adaptation was found to be significantly different with the three comparison signal frequencies used. The greatest amount of measured adaptation occurred with the 4-kHz comparison signal and the least amount with the 0.25 kHz signal. Comparison signal duty cycles of 13 and 23 yielded no difference in the magnitude of measured adaptation.

Comparison of sensitivity for the perception of bodily rotation and the oculogyral illusion

The purpose of this study was to compare the sensitivity of humans to angular accelerati on using the perception of rotation and the perception of the oculogyral illusion as two indicators. Ten men with normal vestibular function were studied in a precision rotation device using a random, forced choice, double staircase method to determine the thresholds. The thresholds for the oculogyral illusion were found to be substantially and significantly lower than thresholds for the perception of rotation. The implications of these findings for an understanding of the oculogyral illusion are discussed.

An Experimental Comparison of the Method of Limits and the Double Staircase-Method

It is a well-known psychophysical fact that the values of thresholds and PSEs depend in part upon the method of measurement.' In the method of constant stimuli, the distribution of values of the stimulus-variable presented to O affects at least measures of central tendency. In the traditional method of limits, the end-point of a run is undoubtedly biased by its starting point. This bias may be particularly serious when only ascending or only descending runs are employed, and when the variability of the endpoints is a function of the parameter under investigation. It may even be present, however, when the two kinds of runs are interspersed, because the amount of the bias probably depends on the distance separating the starting point and the 'true' threshold or PSE. Thus, the method of limits permits E to influence his results, even if only inadvertently. Cornsweet has recently described a variant of the staircase-method (itself a variant of the method of limits) which appears to be largely free of starting point bias.2 In the usual staircase-method, or method of up and down,3 E changes the value of the independent stimulus-variable by the same amount on every trial, but in a direction completely determined by O's response on the just preceding trial: The direction of stimulus-change is reversed when O's response changes, but not otherwise. In the new version, trials from two such staircase-series with different starting points are randomly mixed. This paper presents data from an experiment on the detection of a line-increment, designed to compare Cornsweet's double stair-