# Human Perception Measures for Product Design and Development—A Tutorial to Measurement Methods and Analysis

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

#### 1.1. Perception Studies for Engineers

#### 1.2. Measuring Perception

- detection—detecting, whether there is a stimulus present or not;
- discrimination—detecting, whether two stimuli are different in one or more parameters;
- identification—identify an unknown stimulus from a given set of stimuli;
- scaling—relation of the size of two or more stimuli (or their parameters).

#### 1.3. Structure of This Paper

## 2. Study Design

#### 2.1. Hypothesis

- Light emitting diode (LED) technology allows for control of not only brightness and color of white light consisting of different LED colors, but also has an impact on the color rendering of illuminated objects. The energy consumption of LED lightning can be lowered at the expense of the quality of color rendering. Thompson investigates in [7] whether color rendering quality loss is detectable in peripheral and central vision and proposes an optimized dynamic lightning scheme that reduces energy consumption in peripheral vision without affecting the brightness of the scene.
- New automotive lightning systems promise increased sight and visibility, but potentially increase glare of other road users. In [8], Zydek investigates an improved glare-free lightning system and does not find an increased glare of other road users, but a better sight compared to conventional systems.
- The wide use of touch-based interfaces in consumer electronics triggers new applications with haptic feedback in professional contexts. Personal protection gear such as gloves is more wide-spread in professional applications. The question arises of whether these protectional measures have to be considered in the design of professional haptic interfaces. Two recent studies by Seeger et al. [9] and Hatzfeld et al. [10] investigate absolute and differential thresholds for protective and surgical gloves, respectively, and find different parameters with and without gloves, but not necessarily a need to consider these in the design of haptic interfaces.
- The example in Section 6 investigates the perceptual thresholds of damping in rotary controls in combination with other parameters such as detent or user distraction. It aims to find design parameters for clearly distinguishable controls as well as acceptable tolerance values.

#### 2.1.1. Perception Measures

**Absolute Threshold (AT):**The stimulus magnitude that is required for a positive detection of the stimulus [11]. It is the primary measure for the detection primitive and is sometimes also termed Detection Threshold. If ATs are directly obtained (not calculated from a psychometric function), the detection probability depends on the psychometric procedure used.

**Just Noticeable Difference (JND):**The difference in stimulus magnitude that is required to distinguish two stimuli. It is the primary measure for differentiation capabilities [11,12]. There are two common definitions of the JND:

**Point of Subjective Equality (PSE):**The configuration of two stimuli that are considered as equal by an observer. This measure belongs to the identification primitive [13].

**Just Tolerable Difference (JTD):**The perceived difference between two stimuli that is still tolerable with respect to the intended usage. This measure is also known as Quality JND. Like the above-mentioned PSE, it is a measure of identification, but has much larger importance for the design of technical systems [14,15].

**Power Function Exponent (a):**This measure is derived from Steven’s Power Law (Equation (6)) and is used to describe the correlation between the objective stimulus intensity $\mathrm{\Phi}$ above the absolute threshold ${\mathrm{\Phi}}_{0}$ and the subjective perceived magnitude M:

**Psychological and Psychophysical Measures:**The above-mentioned measures are closely related to the perception primitives given in the introduction (Section 1.2). They are used to quantify the capabilities of sensory and memory processes in the test person.

#### 2.1.2. Typical Experiments for Product Development

**Assessment of a Psychophysical Measure:**In this experiment type, the value of the psychophysical measure itself is of interest. The experiment is conducted with respect to the parameters of the stimulus (for example, absolute threshold of a vibration with respect to frequency) and the most relevant external parameters are considered with respect to typical configurations for the intended application.

**Quantification of Influencing Factors:**In this case, the assessment of external parameters is the focus of the psychophysical test. One wants to know whether and to what extent external parameters have a measurable impact on the human perception.

#### 2.1.3. External Influences

#### 2.2. Classifying Parameters

#### 2.3. Measurement Procedure

#### 2.3.1. Psychometric Methods

**Classic Psychometric Methods**were first developed by Fechner in the end of the 19th century. A straightforward approach is to repeatedly present a number of different stimuli in the expected range of perceivable stimulus intensities, normally about 100 to 200 presentations for a stable result. Based on the perceived stimuli, a psychometric function can be fitted with a number of different fitting algorithms [4,11]. This method is called the Method of Constant Stimuli and one of the methods with the least requirements on a priori knowledge about the psychometric function.

**Non-Parametric Adaptive Methods**are based on a given set of rules for stimulus placement and result calculation. The stimulus placement rule depends on the response of the subject in the course of the experiment. The most prominent non-parametric method is the Staircase Method, which is based on the Method of Limits. The current stimulus level is increased for each negative response of the test person and reduced for each positive response, leading to a threshold with a detection probability of $\mathrm{\Psi}=0.5$. Normally, the stepsizes are fixed. Extensions to this placement rule led to the nowadays frequently used Transformed Up-Down-Staircase [27]. This method changes the stimulus level only after a certain number of consecutive correct or false responses and therefore targets a threshold with a predefined detection probability of ${P}_{\mathrm{\Psi}}\ne 0.5$. The number of selectable detection probabilities, however, is limited and the number of trails is slightly dependent on the chosen detection probability. A staircase-variation by Kaernbach introduces variable step sizes (Weighted Up-Down-Procedure, WUD, [28]) with good performance measures and a freely selectable detection probability.

**Parametric Adaptive Methods**do not focus on the determination of a threshold with a certain detection probability but on the identification of the parameters of a model of the psychometric function. Pre-defined parameter sets (priors) are tested for their ability to account for the responses of the test person. First approaches were made with a Maximum-Likelihood-Estimation of the priors [35], and further developments of this method led to the Updated Maximum Likelihood Procedure (UML) [36]. Another approach was made by Kontsevich and Tyler [37] by developing the $\mathrm{\Psi}$-Method based on conditional probabilities for each of the prior distributions.

#### 2.3.2. Response Paradigms

#### 2.3.3. Selection of a Procedure

- Parametric methods like $\mathrm{\Psi}$ [37] or the Updated-Maximum-Likelihood-Method (UML) [36,42] provide performance benefits for assessing a complete psychometric function. However, they require certain assumptions concerning the psychometric function that may not be available for every kind of experiment [35]. An alternative is the use of the classic Method of Constant Stimuli and finding an acceptable trade-off between accuracy and duration of the experiment.
- For the approximation of the $\mathrm{\Psi}=0.5$ point of the psychometric function, parametric methods are the best choice as well, if an assumption about the form of the function can be made. If that is not the case, one can use the wide-spread adaptive staircase methods, which are easy to implement and only rely on some weak assumptions. An alternative is using approximation methods for stochastic processes such as the ASA procedure [43] or the WUD method [28,38]. These are based on a more complex mathematical basis and can be set to an arbitrary detection probability. Furthermore, they rely on weak assumptions about the psychometric function only and are therefore suitable for experiments with little a priori knowledge about the form and parameters of the psychometric function of the subject.
- For psychometric measures that cannot be described as a parameter of a psychometric function, other types of psychometric procedures must be used. Stevens gives several examples for magnitude estimation tasks [16], in addition to the Method of Adjustment described above.

#### 2.4. Subject Selection

**number of subjects**, no clear answer can be given. In general, one can assume, which the investigation of mainly psychological properties requires more subjects than the investigation of mainly psychophysical properties. A formal deduction of the required number of subjects can be done by considering the allowable $\beta $-error (type II-error, i.e., the probability of retaining a false hypothesis, sometimes also named the power of a study with $\mathrm{Power}=1-\beta $) and the expected effect size of the influencing parameters. This measure is, for example, given by Cohens d in the case of experiments, where means of groups are compared or the population effect size ${\omega}^{2}$ for ANOVA-type analysis. Keppel and Wickens give calculation examples for different types of experiments in [49]. These examples are, however, based on an estimation of the effect size expected, which is difficult to determine without prior knowledge from other experiments.

## 3. Measurement Setup and Errors

**measurement setup**of a perception study has to fit well for the investigated perception parameter. This means, for example, that all parts of the measurement setup exhibit adequate frequency response, rated ranges and sampling rates for the expected values in the experiment. The design and construction of the setup has to be neat to prevent unwanted effects and errors such as errors induced by inadequate electromagnetic shielding. We recommend using fully automated measurement setups to prevent errors induced by the experimenter, i.e., for example by an automated data acquisition instead of reading the results manually from a faltering digital display. The setup has to be documented including all procedures and measurements of systematic and random errors (see below).

**errors**in psychophysical experiments: errors in the perception of the test subject and errors of the measurement setup producing stimuli and measuring signals and answers. The former errors are subject of investigation of the experiment and can be minimized by carefully considering the perception-influencing variables as listed in Section 2.2. This will not yield an error-free measurement, but minimizes the uncertainty of the experimental result.

**Errors of the measurement setup**also diminish the explanatory power of a study and have to be analyzed thoroughly to identify the minimum amount of uncertainty that is associated with the measurement results. An analysis of systematic error propagation should be conducted as well as a calibration documentation of the setup and its components with known input signals and a null signal. Preferably, long time stability, reproducibility and random errors are also analyzed and documented. This can be done by applying the Guide to the Expression of Uncertainty in Measurement (GUM) [51] to the test setup. If an unacceptable large error is derived from this analysis, a larger sample size (i.e., more test subjects) can be used to increase the explanatory power of the experiment.

## 4. Conducting the Test

**pre-test**is recommended to verify assumptions about the psychometric procedures used and the statistical parameters used for the sample size calculation. The researchers working on the study will take part in the pre-test, but at least one person naive to the test should also take part in it to check the understandability of the test instructions.

**Test instructions**should be given in written form to the subjects to avoid unwanted distractions and priming of the subject. Despite this, one should keep in mind to assess the socio-demographic data of the test-subject (if that is necessary for the data analysis) and get a formal consent about the data usage from the test subject. Although intended for medical studies, the Declaration of Helsinki [52] is a de-facto standard for experiments with human test persons and can be used as a gold standard for these procedures.

**interactions with other sensual modalities**should be kept in mind and eventually controlled, for example by ear plugs and masking noise for haptic experiments. For vision, stray lights and insufficient adaption of the observer to the illumination of the setup are undesirable and have to be considered by technical means and adequate adaptation times.

**order of the treatment**for each test subject is supposed to be chosen in such a way, for which learning and habituation as well as systematic effects are minimized. Known methods from Design of Experiments (DoE) [53], i.e., randomization and blocking, as well as Latin Square Designs [54], which will propose a treatment order for each subject, can be used to minimize such effects over the entire sample. This is of utter importance, if the experiment is not fully automated, but relies on interaction with the experimenter. In this case, the experimenter has to be taken into account as a confounding factor.

## 5. Data Analysis

#### 5.1. Checking the Data

#### 5.2. Checking the Hypothesis

#### 5.3. Reporting the Results

## 6. Example: Haptic Perception of Viscous Damping of Rotary Switches

#### 6.1. Measurement Setup and Stimuli

- Knob diameter: ${D}_{1}=20\text{}\mathrm{m}\mathrm{m}$ and ${D}_{2}=38\text{}\mathrm{m}\mathrm{m}.$
- Detent profile: a high-grade detent profile with a maximum torque of 25 $\mathrm{m}$$\mathrm{N}$ $\mathrm{m}$, a slope proportion of 1:5 (rise to fall) and a spatial period of 18${}^{\circ}$ can be superimposed on the stimulus. Experimental conditions are activated detent or no detent.
- Distraction: thresholds are either determined as a primary task (no distraction condition), or as a secondary task to a standardized Lane Change Test (LCT) [63] (distraction condition).

#### 6.2. Subjects

#### 6.3. Measurement Procedure

#### 6.4. Results

#### 6.5. Discussion

## 7. Conclusions

## Acknowledgments

## Author Contributions

## Conflicts of Interest

## References

- Kern, T.A.; Hatzfeld, C. The User’s Role in Haptic System Design. In Engineering Haptic Devices, 2nd ed.; Hatzfeld, C., Kern, T.A., Eds.; Springer: London, UK, 2014. [Google Scholar]
- Weber, E.H. Tastsinn und Gemeingefühl; Engelmann: Leipzig, 1905. [Google Scholar]
- Gall, S.; Beins, B.; Feldman, J. Psychophysics. In The Gale Encyclopedia of Psychology; Gale: Detroit, Michigan, 2001. [Google Scholar]
- Wichmann, F.; Hill, N. The psychometric function: I. Fitting, Sampling, and Goodness of Fit. Percept. Psychophys.
**2001**, 63, 1293–1313. [Google Scholar] [CrossRef] [PubMed] - Kühner, M.; Wild, J.; Bubb, H.; Bengler, K.; Schneider, J. Haptic Perception of Viscous Friction of Rotary Switches. In Proceedings of the 2011 IEEE World Haptics Conference (WHC), Istanbul, Turkey, 21–24 June 2011; pp. 587–591. [Google Scholar]
- Kühner, M. Haptische Unterscheidbarkeit Mechanischer Parameter bei Rotatorischen Bedienelementen. Ph.D. Thesis, Technische Universität München, Munich, Germany, 2014. [Google Scholar]
- Do Thompson, M.R. Psychophysical Evaluations of Modulated Color Rendering for Energy Performance of LED-Based Architectural Lighting. Ph.D. Thesis, Massachusetts Institute of Technology, Cambridge, MA, USA, 2007. [Google Scholar]
- Zydek, B. Blendungsbewertung von Kraftfahrzeugscheinwerfern unter Dynamischen Bedingungen. Ph.D. Thesis, Technische Universität Darmstadt, Darmstadt, Germany, 2014. [Google Scholar]
- Seeger, M.; Stein, T.; Schmidt, L. Vibrotaktile Wahrnehmung bei der Verwendung von Handschuhen. In Mensch und Computer 2017—Tagungsband; Gesellschaft für Informatik: Regensburg, Germany, 2017. [Google Scholar]
- Hatzfeld, C.; Dorsch, S.; Neupert, C.; Kupnik, M. Influence of surgical gloves on haptic perception thresholds. Int. J. Med. Robot. Comput. Assist. Surg.
**2017**. [Google Scholar] [CrossRef] [PubMed] - Prins, N.; Kingdom, F.A.A. Psychophysics: A Practical Introduction; Academic Press: Maryland Heights, MO, USA, 2010. [Google Scholar]
- Macmillan, N.; Creelman, C. Detection Theory: A User’s Guide; Lawrence Erlbaum: London, UK, 2005. [Google Scholar]
- Colman, A. A Dictionary of Psychology, 3rd ed.; Oxford University Press: Oxford, UK, 2009. [Google Scholar]
- Conner, M.; Booth, D.; Clifton, V.; Griffiths, R. Individualized optimization of the salt content of white bread for acceptability. J. Food Sci.
**1988**, 53, 549–554. [Google Scholar] [CrossRef] - International Organization for Standardization (ISO). ISO 20462-3: Photography—Psychophysical Experimental Methods for Estimating Image Quality; ISO: Geneva, Switzerland, 2012. [Google Scholar]
- Stevens, S.S. Psychophysics; Transaction Books: Piscataway, NJ, USA, 1975. [Google Scholar]
- Theeuwes, J.; Alferdinck, J.W.; Perel, M. Relation between glare and driving performance. Hum. Factors J. Hum. Factors Ergonom. Soc.
**2002**, 44, 95–107. [Google Scholar] [CrossRef] [PubMed] - Hatzfeld, C. Haptics as an Interaction Modality. In Engineering Haptic Devices, 2nd ed.; Hatzfeld, C., Kern, T.A., Eds.; Springer: London, UK, 2014. [Google Scholar]
- Zwicker, E.; Fastl, H. Psychoacoustics; Springer: Heidelberg, Germany, 1999. [Google Scholar]
- Drasdo, N. Vision Research: A Practical Guide to Laboratory Methods. Brain
**1999**, 122, 2000–2002. [Google Scholar] [CrossRef] - Peters, R.; Hackeman, E.; Goldreich, D. Diminutive Digits Discern Delicate Details: Fingertip Size and the Sex Difference in Tactile Spatial Acuity. J. Neurosci.
**2009**, 29, 15756–15761. [Google Scholar] [CrossRef] [PubMed] - Verrillo, R.; Bolanowski, S. The Effects of Skin Temperature on the Psychophysical Responses to Vibration on Glabrous and Hairy Skin. J. Acoust. Soc. Am.
**1986**, 80, 528–532. [Google Scholar] [CrossRef] [PubMed] - Gescheider, G.A.; Bolanowski, S.J.; Hall, K.L.; Hoffman, K.E.; Verrillo, R.T. The Effects of Aging on Information-Processing Channels in the Sense of Touch: I. Absolute Sensitivity. Somatosens. Mot. Res.
**1994**, 11, 345–357. [Google Scholar] [CrossRef] [PubMed] - Fechner, G.T. Elemente der Psychophysik; Breitkopf und Härtel: Leipzig, Germany, 1860. [Google Scholar]
- Green, D.M.; Swets, J.A. Signal Detection Theory and Psychophysics; Wiley: New York, NY, USA, 1966. [Google Scholar]
- Wickens, T.D. Elementary Signal Detection Theory; Oxford University Press: Oxford, UK, 2002. [Google Scholar]
- Levitt, H. Transformed Up-Down Methods in Psychoacoustics. J. Acoust.Soc. Am.
**1971**, 49, 467–477. [Google Scholar] [CrossRef] - Kaernbach, C. Simple adaptive testing with the weighted up-down method. Atten. Percept. Psychophys.
**1991**, 49, 227–229. [Google Scholar] [CrossRef] - Taylor, M.; Creelman, C. PEST: Efficient Estimates on Probability Functions. J. Acoust. Soc. Am.
**1967**, 41, 782–787. [Google Scholar] [CrossRef] - Kesten, H. Accelerated Stochastic Approximation. Ann. Math. Stat.
**1958**, 29, 41–59. [Google Scholar] [CrossRef] - Tyrrell, R.A.; Owens, D.A. A rapid technique to assess the resting states of the eyes and other threshold phenomena: the modified binary search (MOBS). Behav. Res. Methods Instrum. Comput.
**1988**, 20, 137–141. [Google Scholar] [CrossRef] - Leek, M.R. Adaptive Procedures in Psychophysical Research. Percept. Psychophys.
**2001**, 63, 1279–1292. [Google Scholar] [CrossRef] [PubMed] - Hatzfeld, C.; Kupnik, M.; Werthschutzky, R. Performance simulation of unforced choice paradigms in parametric psychometric procedures. In Proceedings of the 2015 IEEE World Haptics Conference (WHC), Evanston, IL, USA, 22–26 June 2015; pp. 475–481. [Google Scholar]
- Otto, S.; Weinzierl, S. Comparative Simulations of Adaptive Psychometric Procedures; Jahrestagung der Deutschen Gesellschaft für Akustik: Berlin, Germany, 2009; pp. 1276–1279. [Google Scholar]
- Harvey, L.O. Efficient Estimation of Sensory Thresholds. Behav. Res. Methods
**1986**, 18, 623–632. [Google Scholar] [CrossRef] - Shen, Y.; Richards, V.M. A maximum-likelihood procedure for estimating psychometric functions: Thresholds, slopes, and lapses of attention. J. Acoust. Soc. Am.
**2012**, 132, 957–967. [Google Scholar] [CrossRef] [PubMed] - Kontsevich, L.; Tyler, C. Bayesian Adaptive Estimation of Psychometric Slope and Threshold. Vis. Res.
**1999**, 39, 2729–2737. [Google Scholar] [CrossRef] - Kaernbach, C. Adaptive Threshold Estimation with Unforced-Choice Tasks. Percept. Psychophys.
**2001**, 63, 1377–1388. [Google Scholar] [CrossRef] [PubMed] - Prins, N.; Kingdom, F.A.A. Palamedes: Matlab Routines for Analyzing Psychophysical Data. Available online: http://www.palamedestoolbox.org/ (accessed on 28 July 2017).
- Karmali, F.; Chaudhuri, S.E.; Yi, Y.; Merfeld, D.M. Determining thresholds using adaptive procedures and psychometric fits: Evaluating efficiency using theory, simulations, and human experiments. Exp. Brain Res.
**2016**. [Google Scholar] [CrossRef] [PubMed] - Madigan, R.; Williams, D. Maximum-Likelihood Psychometric Procedures in Two-Alternative Forced-Choice: Evaluation and Recommendations. Atten. Percept. Psychophys.
**1987**, 42, 240–249. [Google Scholar] [CrossRef] - Green, D. Stimulus selection in adaptive psychophysical procedures. J. Acoust. Soc. Am.
**1990**, 87, 2662–2674. [Google Scholar] [CrossRef] [PubMed] - Anderson, A.J.; Johnson, C.A. Comparison of the ASA, MOBS, and ZEST threshold methods. Vis. Res.
**2006**, 46, 2403–2411. [Google Scholar] [CrossRef] [PubMed] - Jäkel, F.; Wichmann, F.A. Spatial four-alternative forced-choice method is the preferred psychophysical method for naïve observers. J. Vis.
**2006**, 6, 1307–1322. [Google Scholar] [CrossRef] [PubMed] - Kollmeier, B.; Gilkey, R.H.; Sieben, U.K. Adaptive Staircase Techniques in Psychoacoustics: A Comparison of Human Data and a Mathematical Model. J. Acoust. Soc. Am.
**1988**, 83, 1852–1862. [Google Scholar] [CrossRef] [PubMed] - Schlauch, R.; Rose, R. Two-, Three-, and Four-Interval Forced-Choice Staircase Procedures: Estimator Bias and Efficiency. J. Acoust. Soc. Am.
**1990**, 88, 732. [Google Scholar] [CrossRef] [PubMed] - Cascio, C.; McGlone, F.; Folger, S.; Tannan, V.; Baranek, G.; Pelphrey, K.A.; Essick, G. Tactile Perception in Adults with Autism: a Multidimensional Psychophysical Study. J. Autism Dev. Disord.
**2008**, 38, 127–137. [Google Scholar] [CrossRef] [PubMed] - Grunwald, M.; Ettrich, C.; Krause, W.; Assmann, B.; Dähne, A.; Weiss, T.; Gertz, H.J. Haptic perception in anorexia nervosa before and after weight gain. J. Clin. Exp. Neuropsychol.
**2001**, 23, 520–529. [Google Scholar] [CrossRef] [PubMed] - Keppel, G. Design and Analysis: A Researcher’s Handbook; Pearson Education: Old Tappan, NJ, USA, 1991. [Google Scholar]
- Faul, F.; Erdfelder, E.; Lang, A.G.; Buchner, A. G* Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav. Res. Methods
**2007**, 39, 175–191. [Google Scholar] [CrossRef] [PubMed] - International Organization for Standardization (ISO). ISO/IEC Guide 98-3: Uncertainty of Measurement—Part 3: Guide to the Expression of Uncertainty in Measurement; ISO: Geneva, Switzerland, 2008. [Google Scholar]
- Declaration of Helsinki—Ethical Principles for Medical Research Involving Human Subjects. Technical Report. World Medical Association, 2013. Available online: http://www.wma.net/en/30publications/10policies/b3/ (accessed on 28 July 2017).
- Montgomery, D.C. Design and Analysis of Experiments; John Wiley & Sons: Hoboken, NJ, USA, 2009. [Google Scholar]
- Ai, M.; Li, K.; Liu, S.; Lin, D.K. Balanced incomplete Latin square designs. J. Stat. Plan. Infer.
**2013**, 143, 1575–1582. [Google Scholar] [CrossRef] - Christensen, R. Testing Fisher, Neyman, Pearson, and Bayes. Am. Stat.
**2005**, 59, 121–126. [Google Scholar] [CrossRef] - Razali, N.M.; Wah, Y.B. Power comparisons of shapiro-wilk, kolmogorov-smirnov, lilliefors and anderson-darling tests. J. Stat. Model. Anal.
**2011**, 2, 21–33. [Google Scholar] - Wallenstein, S.; Zucker, C.L.; Fleiss, J.L. Some statistical methods useful in circulation research. Circ. Res.
**1980**, 47, 1–9. [Google Scholar] [CrossRef] [PubMed] - Field, A. Discovering Statistics Using IBM SPSS Statistics, 4th ed.; Sage Publications: New York, NY, USA, 2014. [Google Scholar]
- Levine, T.R.; Hullett, C.R. Eta squared, partial eta squared, and misreporting of effect size in communication research. Hum. Commun. Res.
**2002**, 28, 612–625. [Google Scholar] [CrossRef] - Olejnik, S.; Algina, J. Generalized eta and omega squared statistics: Measures of effect size for some common research designs. Psychol. Methods
**2003**, 8, 434–447. [Google Scholar] [CrossRef] [PubMed] - Nitsch, V.; Färber, B. A Meta-Analysis of the Effects of Haptic Interfaces on Task Performance with Teleoperation Systems. IEEE Trans. Hapt.
**2012**, 6, 387–398. [Google Scholar] [CrossRef] [PubMed] - Hogan, N. Impedance Control: An Approach to Manipulation. ASME J. Dyn. Syst. Meas. Control
**1985**, 107, 1–24. [Google Scholar] [CrossRef] - International Organization for Standardization (ISO). ISO 26022: Road Vehicles—Ergonomic Aspects of Transport Information and Control Systems—Simulated Lane Change Test to assess in-Vehicle Secondary Task Demand; ISO: Geneva, Switzerland, 2010. [Google Scholar]
- Kühner, M.; Bubb, H.; Bengler, K.; Wild, J. Adaptive Verfahren in der Psychophysik. In Ergonomie; Lehrstuhl für Ergonomie, Technische Universität München: München, Germany, 2012; p. 26. [Google Scholar]
- Cohen, J. Statistical Power Analysis for the Behavioral Sciences; Academic Press: Cambridge, MA, USA, 1988. [Google Scholar]

**Figure 1.**Sample psychometric function $\mathrm{\Psi}(\mathrm{\Phi})$ with location parameter $\alpha =8$, sensitivity parameter $\beta =1$, corrected guess rate $\gamma =0.15$ and lapse rate $\lambda =0.075$. A stimulus with intensity $\mathrm{\Phi}=\alpha =8$ is detected in 50% of all trials. For discrimination tasks, the relation between test and reference stimulus is often used as stimulus $\mathrm{\Phi}$.

**Figure 2.**Examples of psychometric methods: Transformed Up-Down-Staircase with 1up-3down progression rule (

**top**) and $\mathrm{\Psi}$ (

**bottom**). Graphs are taken from a simulation and show the tested stimulus intensity for each trail. Red squares denote negative, green dots denote positive responses of the test persons to the stimulus. The dotted line shows the calculated threshold, and reversals of the staircase method are denoted by larger circles.

**Figure 3.**Calculation of sample sizes using G*Power 3.19 for a two-factor, within-subject design with four treatment groups and a single dependent variable.

**Figure 4.**Seatbox with steering wheel (

**a**) and screenshot of the lane change test (

**b**). Monitor, headphones for acoustic shielding and haptic simulator are not shown in (

**a**).

**Figure 5.**Boxplots of obtained absolute thresholds. Boxplots denote median (thick red vertical line) and interquartile range (IQR) between 0.25 and 0.75 quantile (blue box). Notches denote the confidence intervals ($\alpha =0.05$) of the median values. Outliers (circles) are defined as data points with a distance of more than 1.5 IQR from 0.25 or 0.75 quantile, respectively, as indicated by the horizontal dashed lines. Stimulus conditions are coded with diameter (D1/D2), distraction (LCT/-) and detent condition (D/-). Boxplots show an increase in data span for distraction as well as detent condition. Non-overlapping notches of the central boxes hint at significant median differences of the data sets.

**Figure 6.**Effect of distraction on detection thresholds for damping with respect to knob diameter. Differences are not significant.

**Figure 7.**Effect of knob diameter on detection thresholds with respect to distraction and detent. Differences are significantly different from zero and are correlated to knob diameter.

**Figure 8.**Effect of detent condition on detection thresholds with respect to knob diameter. Differences are significantly different from zero.

Type of Variable | Haptics | Vision |
---|---|---|

Dependent variable | Psychophysical construct in terms of JND, PSD etc. | |

Independent variable | Stimulus parameter (frequency, intensity etc.), contact area [21], contact force, masking stimuli | Stimulus parameter (size, intensity), time of adaption |

Controllable variable | Skin moisture, skin temperature [22], test person’s age [23], test person’s sex | Amblyopia, spectral power distribution (SPD) of the stimulus, adaption field |

Confounding variables | Fatigue, experience of test person, other modalities, change of experimenter, non-thought-of variables |

**Table 2.**Absolute detection thresholds for damping in different conditions with knob diameter (D1/D2), distraction (LCT/-), and detent (D/-). Values are given in Section 6.1.

Condition | Absolute Threshold (mN m s) | ||||||
---|---|---|---|---|---|---|---|

Min. | ${\mathit{Q}}_{\mathbf{0.25}}$ | Median | ${\mathit{Q}}_{\mathbf{0.75}}$ | Max. | $\widehat{\mathit{\mu}}$ | $\widehat{\mathit{\sigma}}$ | |

D1 | 0.015 | 0.046 | 0.105 | 0.164 | 0.455 | 0.148 | 0.135 |

D1.LCT | 0.005 | 0.07 | 0.135 | 0.225 | 0.305 | 0.149 | 0.087 |

D1.LCT.D | 0.065 | 0.255 | 0.355 | 0.495 | 0.87 | 0.386 | 0.216 |

D2 | 0.035 | 0.135 | 0.235 | 0.35 | 0.455 | 0.245 | 0.128 |

D2.LCT | 0.025 | 0.161 | 0.298 | 0.439 | 0.745 | 0.312 | 0.184 |

D2.LCT.D | 0.065 | 0.386 | 0.54 | 0.703 | 1.395 | 0.598 | 0.349 |

**Table 3.**Weber fractions for damping in different conditions with knob diameter (D1/D2), distraction (LCT/-), and detent (D/-). Values are given in Section 6.1.

Condition | Weber Fraction (%) | ||||||
---|---|---|---|---|---|---|---|

Min. | ${\mathit{Q}}_{\mathbf{0.25}}$ | Median | ${\mathit{Q}}_{\mathbf{0.75}}$ | Max. | $\widehat{\mathit{\mu}}$ | $\widehat{\mathit{\sigma}}$ | |

D1 | 0.40 | 10.00 | 17.90 | 30.85 | 50.40 | 19.79 | 13.95 |

D1.LCT | 0.85 | 17.10 | 22.30 | 31.56 | 44.15 | 22.84 | 10.72 |

D1.LCT.D | 1.25 | 22.20 | 29.80 | 38.11 | 52.90 | 28.86 | 14.57 |

D2 | 1.65 | 15.00 | 18.75 | 27.90 | 48.75 | 20.91 | 10.96 |

D2.LCT | 1.25 | 13.96 | 21.88 | 27.40 | 48.75 | 22.00 | 11.63 |

D2.LCT.D | 1.65 | 20.01 | 29.58 | 41.03 | 70.00 | 30.98 | 17.36 |

Effect | Significance | Mean Effect Size ${\mathit{d}}^{\mathbf{\prime}\mathbf{\prime}}$ |
---|---|---|

Absolute Threshold | ||

Distraction | no | 0.172 |

Detent | yes | 1.126 |

Knob diameter | yes | 0.796 |

Differential threshold | ||

Distraction | no | 0.164 |

Detent | yes | 0.571 |

Knob diameter | no | 0.098 |

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Hatzfeld, C.; Kühner, M.; Söllner, S.; Khanh, T.Q.; Kupnik, M.
Human Perception Measures for Product Design and Development—A Tutorial to Measurement Methods and Analysis. *Multimodal Technol. Interact.* **2017**, *1*, 28.
https://doi.org/10.3390/mti1040028

**AMA Style**

Hatzfeld C, Kühner M, Söllner S, Khanh TQ, Kupnik M.
Human Perception Measures for Product Design and Development—A Tutorial to Measurement Methods and Analysis. *Multimodal Technologies and Interaction*. 2017; 1(4):28.
https://doi.org/10.3390/mti1040028

**Chicago/Turabian Style**

Hatzfeld, Christian, Manuel Kühner, Stefan Söllner, Tran Quoc Khanh, and Mario Kupnik.
2017. "Human Perception Measures for Product Design and Development—A Tutorial to Measurement Methods and Analysis" *Multimodal Technologies and Interaction* 1, no. 4: 28.
https://doi.org/10.3390/mti1040028