The second part of the book discusses the different methods used for electrodermal recording. As mentioned in the introduction to Part 1, the observation of electrodermal phenomena is possible with relatively simple equipment, resulting in a variety of methodologies.


Sweat Gland Skin Resistance Amplitude Criterion Electrodermal Activity Sweat Gland Activity 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 28.
    Although the respective EDL is built up of physical units with ratio scales, it is sometimes recommended that the EDL signal — like most psychophysiological variables — be treated as based only on interval scales (Levey, 1980; Stemmler, 1984), which would naturally also affect the performance of transformation (Sect. 2.3.3). However, EDR amplitudes obtained by AC-coupled amplification cannot be treated as based on a ratio scale anyhow.Google Scholar
  2. 32.
    According to Venables and Christie (1980), no differences of potentials between abraded and non-abraded forearm sites were observable with children, so when children are used as subjects, pretreatment can be omitted.Google Scholar
  3. 33.
    The relative error due to seepage is dependent on electrode diameter. 1 mm of seepage increases the contact area to 2.25 times its original size with 4 mm diameter miniature electrodes, but to only 1.13 times with the use of 1 cm diameter electrodes (Venables & Christie, 1980).Google Scholar
  4. 36.
    The Beckman biopotential electrodes (see next footnote) show bias potentials of less than 250 pV and polarization potentials of less than 5 pV (Venables & Christie, 1980). Ag/AgC1 electrodes can also be homemade, albeit unsintered (cf. Venables & Christie, 1973, p. 107), and good results are obtainable, but the process is very expensive, pure silver (99.99%) being necessary. Usually commercially available Ag/AgCI electrodes are preferred.Google Scholar
  5. 38.
    The evidence presented is not very compelling since the air-dry method was best overall (see Tassi-nary et al., 1990, Fig. 2, p. 239). Another problem with the proposed storage method is one from the present author’s own experience. Long-term immersion in an NaCI solution led to a separation of the wire from the chamber at the chamber-wire junction. This problem with wire separation and the trouble and expense of the proposed method still leaves the air-dry method as most recommendable.Google Scholar
  6. 40.
    Sensor Medics recommends removing the deposit with dilute ammonium hydroxide. A five to one dilution with distilled water has been used with success (personal communication, A. Vincent, July 1990).Google Scholar
  7. 41.
    This is also true of the electrode cream “Synapse” made by Beckman Instruments which has been frequently used in EDA measurements. An analysis of this and other gels was made by Zipp, Henne-mann, Grunwald, and Rohmert (1980), who found in “Synapse” a significant quantity of K and CI ions in addition to Na ions. Grey and Smith (1984) reported that the Beckman paste has a NaC1 concentration of 4.1 moll, and contains glycerol, gum tragacanth, and.5% benzyl alcohol (as preservative). An isotonic cream from Heilige, Freiburg (see Footnote 1 in Boucsein and Hoffmann, 1979) is no longer available.Google Scholar
  8. 42.
    Grey and Smith (1984) used a homemade gel of.05 molar NaC1 solution in methyl cellulose. However, they did not provide the correct formula of the ingredients they used (Clements, 1989).Google Scholar
  9. 43.
    Unibase is produced in the U.S. by Warner Chilcott Laboratories, Morris Plain, NJ. Grey and Smith (1984) have published the following ingredients of Unibase: cetyl and stearyl alcohols, soft paraffin, glycerol, and, as preservatives,.0015% propyl hydroxy-benzoate, sodium citrate, and sodium lauryl sulphate. The relative quantities are not provided. The water content is 63,4%. According to Grey and Smith, Unibase contains.028 moI/1 Na ions. The author has had this recipe chemically analyzed; the results were.07 moly Na and.045 mol/1 Cl, the increased sodium being due to the Unibase itself (see previous footnote). The cream is free from K and Ca ions (less than.01 g/kg) and has a nearly neutral pH value of 6.5. The analysis was performed by B. Neidhart, Institute for Industrial Physiology at the University of Dortmund, Germany.Google Scholar
  10. 44.
    Since Lykken and Venables (1971) recommended skin conductance as being the appropriate unit for EDA measurement, constant current recording went out of use in most psychophysiological laboratories. However, because of the lower amplifier gain required, constant current methods are frequently preferred in field applications (Sect. 2.1.1 & 2.6.2).Google Scholar
  11. 45.
    Though Venables and Christie (1980, p. 40) argue for the application of.5 V across each active site, following a suggestion by Edelberg (1967), their circuitry depicted in Figure 31 provides a total of.5 V.Google Scholar
  12. 53.
    Thom (1988) uses a criterion of 10% instead, because the application of the 1% criterion is difficult if numerous electrodermal fluctuations appear.Google Scholar
  13. 56.
    Refined mathematical modelling of EDR curves has been performed by Hunt (1977), who developed an equation based on overlapping Gaussian distributions to fit the course of SRRs, and by Schneider (1987). Schneider fitted a three-compartment model to the recorded SC curve (personal communication) which includes the physical properties of the duct filling, the active membrane response in the duct walls, and the corneal hydration (Sect. 1.4.2). Schneider could show that a typical SCR can be modelled by assuming a roughly triangular input signal and choosing as an impulse response a sum of two exponentials with time constants of approximately 2 and 20 sec, respectively.Google Scholar
  14. 59.
    This has been demonstrated by Boucsein et al. (1984a) using both recording methods in parallel during the application of 2-sec white noise stimuli with intensities between 60 and 110 dB.Google Scholar
  15. 63.
    One interesting result should be mentioned here: Christie and Venables (1971) found a correlation between the BSPL (Sect. and the T-wave amplitude (TWA) in the EKG, with 21 male subjects lying down (r= 70) as well as with 15 subjects in a sitting position (r=-61). The authors suggest the extracellular potassium ion concentration as being the cause for both the TWA and the negativity of the BSPL. A more recent paper by Furedy and Heslegrave (1983) suggests that the TWA is an index of excitatory sympathetic activity. Since it is known that the EDA is a valid index of sympathetic excitation, a high correlation between TWA and EDA is expected.Google Scholar
  16. 65.
    Measuring sweat secretion by quantification of skin moisture is provided by gravi-metric methods, where a humidity sensitive film is attached to the skin, or by colori-metric techniques (e.g., fingerprints). With the aid of a magnifying glass, the number of active (i.e., completely filled) sweat glands per skin area can be obtained with the latter method (Malmo, 1965). Thomas and Korr (1957), studying six subjects, found within-subject correlations ranging from.44 to.96 between sweat gland counts, which were recorded photographically, and the SCL.fi9 Johnson and Landon (1965), in their study described in Section, used a technique developed by Sutarman and Thomson (1952), in which a plastic ink impression of the skin is transferred to Scotch tape and studied microscopically with a magnification of 25.7° Their intraindividual correlations between SCL and the number of active glands were between.29 and.79, thus, not as high on average as those reported by Thomas and Korr (1957). Köhler, Vögele, and Weber (1989) used a fixation solution containing polyvinyl formaldehyde, removed with a Scotch tape strip from the finger and used to count the active sweat glands microscopically. With 20 subjects, they found a correlation between the occurrence of palmar sweating and the SCL change during a stress film of.71. All other correlations were not significant.Google Scholar
  17. 84.
    For discussion of experimenter’s ethnic group on subject’s physiological reactions see Venables and Christie (1973).Google Scholar
  18. 93.
    A review of reliabilities of different EDA parameters, including various investigations, is given by Freixa i Baqué (1982).Google Scholar
  19. 100.
    lacono, Lykken, Haroian, Peloquin, Valentine, and Tuason (1984a), in their study described in Section, found a one-year retest reliability of the maximum SCR amp. during series of tones of r=.68 in 23 normal subjects.Google Scholar
  20. 111.
    A thorough discussion of such corrections which are based either on the use of transformations or on regression techniques is found in Levey (1980, p. 619ff.).Google Scholar
  21. 114.
    This discussion has been used by Catania et al. (1980) to explain the dependency of age-related differences in electrodermal reactivity upon the method of measurement (Sect. Scholar
  22. 115.
    The points raised by Edelberg (1967, p. 25f.) on technical control of the current density with constant voltage systems are not discussed further here since they are taken care of in modem equipment (Sect. 2.2.4).Google Scholar
  23. 118.
    In a study where 20 subjects were presented with 10 tones of 50 dB each and additionally, white quadrangles as stimuli. With the constant current measurements, Barry (1981) used polarizable electrodes and nonisotonic cream, while the constant voltage measurements were made using standard methodology.Google Scholar

Copyright information

© Springer Science+Business Media New York 1992

Authors and Affiliations

  • Wolfram Boucsein
    • 1
  1. 1.University of WuppertalWuppertalGermany

Personalised recommendations