Gunnar Borg

Professor emeritus

Visa sidan på svenska
Works at Department of Psychology
Visiting address Frescati hagväg 14
Postal address Psykologiska institutionen 106 91 Stockholm


A selection from Stockholm University publication database
  • 2014. Gunnar Borg, Elisabet Borg. Fechner Day 2014, 10-10

    There are many demands on a good pain-scale. It should be possible to: use for all kinds of pain; determine direct levels of intensity over the total range; treat responses with statistical methods, preferably parametric statistics; study degrees of changes with stimulus intensity, medication and time; make interindividual, intermodal and psychophysiological comparisons; avoid ceiling and floor effects; make estimations and also productions; determine psychophysical S-R-functions, possible to describe with a general equations as, e.g., R=a+c(S-b)^n, where a is the basic “noise” at rest  (or the absolute threshold), and b is the starting point of the function; make two-way communication; handle round off tendencies; use internationally. To meet these demands the scale must be constructed according to basic psychophysical and linguistic knowledge, and tested in relevant experiments. To cover the total subjective range there is a need of a number variation from 0 to 50 or a little more, about 26. Several anchors should be used that people understand very well, and that are placed correctly. Most existing scales do not fulfill these demands. A common drawback is that there is a too limited range, or a maximal endpoint defined as “Highest (or Worst) Imaginable”, which is not a schematized conception and problematic for interindividual comparisons. Examples are the Visual Analogue Scale (VAS), and is the "Labeled Magnitude Scale" (LMS) for oral sensation. On the LMS verbal anchors are placed to give ratio data, but “Strong” is 34.7, “Very strong” 52,5. For general usage, e.g., in two-way communication for prescription of exercise, this is not good, nor does the scales facilitate predictions of max-levels from sub-max estimations. The best scales are the Borg CR Scales® (CR10 and CR100). In these scales quantitative semantics is used by applying ratio scaling to determine interpretation, meaning position in the range for congruence between anchors (labels) and numbers, and preciseness meaning interindividual agreement. It is especially important that the anchors for Zero and Maximal refer to schematized conceptions. A maximal magnitude is defined as a maximal perceived exertion and effort, for example a maximal heaviness. These ideas have been presented during several ISP meetings by G. Borg, last time in Freiburg 2013. The CR10 has been used in many studies, e.g., during tests of functional capacity and chest pain, and muscular-skeletal pain. The CR100 scale has, however, a greater potential as a general scale making possible determinations of most kinds of perceptual magnitudes. An advantage to the CR10 is that decimals need not be used and that the dynamic range is bigger and more in accordance with the psychophysical demands. The extra constants in the power function can then better reflect the true sensory processes.

  • 2013. Elisabet Borg, Gunnar Borg. Applied Ergonomics 44 (5), 835-840

    Level-anchored ratio scaling, such as the Borg CR10 scale (R) and the Borg CR100 scale (R), uses verbal anchors in congruence with numbers to give ratio data together with natural levels of intensity. This presupposes that the anchors possess natural positions in the subjective dynamic range and also numerical inter-relations. In an experiment, subjects had to produce a force of handgrip corresponding to their conception of Strong, followed by a Maximal performance. By using the previously found relationship between Strong and Maximal of 1:2 together with knowledge of the exponent in the power S-R-function (R = c x S-n) for grip strength, n = 1.8, predictions of individual maximal performances were obtained. The predicted values correlated 0.76 with, and deviated only 3% (ns) from, actual maximal performances of grip strength. This result as previously also found for aerobic capacity gives a strong support for the use of verbal anchors, so common in category scaling, also in ratio scaling and that the Borg CR-scales fulfill the requirements for ratio scales. For estimation of muscular strength, such as grip strength, this present study points to the value of using submaximal determinations as a compliment to maximal performances (e.g., to obtain measures of functional capacity). The results also support the increasingly common use of the CR-methodology in other ergonomic settings concerning suitable design of tools and equipment.

  • 2013. Elisabet Borg, Gunnar Borg. Fechner Day 2013, 97-97

    Pain is one of the most common symtoms reported clinically. Apart from the Visual Analogue scale, several rating scales are used varying in degree of interpretability and suitability for various types of pain. One scale often used, especially for children, is the Wong–Baker FACES Pain Rating Scale where pain is expressed in six drawn faces varying in expression from (1) a smiling face denoting “no hurt” to (6) a crying face for “hurts worst”. Two experiments were carried out. Firstly, 12 university professors of psychology (8 men and 4 women, 50–79 yrs) answered where (in what face) they judged that pain with certainty started. Three answered that pain started in the sixth face. Two said that no face showed pain and the rest said that the dimension shown was degree of happiness-sadness. However, if the intensity of pain was estimated as if using cross-modality matching (sadness to pain), three said face no. 4, one no. 4-5, one no. 4,5, or 6, and two said face number 5 (median = 4.5). Secondly, sixteen university students (4 men and 12 women, mean age = 27.2, SD = 7.4 yrs) answered the same question as above and then also used the Borg CR100 (centiMax) Scale®  , a general 0—100 intensity Category-Ratio scale for most kinds of subjective measurement3 , to scale the pain intensity expressed in each of the six faces (presented twice in a randomized order in a Powerpoint presentation). Pain was “with certainty” judged to start at the fifth face, and several participants scaled the first two faces as “zero pain”. A continuous progression of pain intensity for the six faces was on the average obtained with the CR100 scale: medians = 0.0, 0.5, 11, 31, 48, 72 centiMax. Thus, expressed with the verbal labels on the CR100 scale, the first two faces were below “Minimal”, which indicates that these faces were not judged to show any pain, the third face was just above “Weak” (13), the fourth face just above “Moderate” (25), the 5th face just below “Strong” (50) and the 6th face was just below “Very strong” (70). The conclusion was that the faces only with hesitation can be used to estimate pain. The last face was not judged to show more than a very strong pain, thus causing a restriction of range and a ceiling effect. Aproblem with these kinds of scales is poor congruence between pictures, verbal labels and numbers.

  • 2012. Elisabet Borg, Gunnar Borg. Proceedings of Fechner Day 28, 143-146

    In some sports, as for example in diving, performance is measured as a subjectively evaluated artistic gestalt. The purpose of this study was to compare the traditional scale used in competitive diving with the Borg CR100 scale®, a scale where categorical expressions are placed where they perceptually belong on a ratio scale (e.g., G. Borg and E. Borg, 2001). Two internationally recognized Swedish judges volunteered as subjects and judged a sample of 45 videotaped dives, both with the traditional scale and with the CR scale. The results show that the Borg CR100 scale® worked at least equally well as the traditional scale, even though there might have been some tendency for translation between scales.

  • 2011. John P. Buckley, Gunnar A. V. Borg. Applied Physiology, Nutrition and Metabolism 36 (5), 682-692

    This study is the first to apply Borg's psychophysical equation to measuring responses to strength training with weights machines. Theoretical constructs of Borg's scales were assessed in younger and older adults to estimate the appropriate load and number of repetitions required to meet recommended practice guidelines. A younger group (YG; 20 males, 20 females; aged 19-38 years) and older group (OG; 13 males, 13 females; aged 50-75 years) participated in 3 experiments. Experiment 1: YG performed 2-repetitions of incremented loads during triceps-elbow extensions and knee extensions to level 7 on Borg's CR10 Scale. Experiment 2: YG (n = 16) then performed 12-repetitions at the loads from experiement 1 that elicited CR10 ratings 1.5, 3.0, and 5.0. Experiment 3: OG performed 15-repetitions of lat-pull and leg press at 15-repetition maximum (RM) load. In experiments 2 and 3, CR10 or Borg RPE were measured every 2 repetitions. Experiment 1 revealed classic psychophysical response growth exponents between 1.1 and 1.8, which were greater in arms than legs (p < 0.001) and in females (p < 0.001). Theoretical estimates of 1RM were derived from the growth curves for the weights eliciting CR10 ratings of 1.5, 3, and 5. CR10 ratings of 3 to 6 fell within estimates of 40%-70% 1RM. Experiments 2 and 3 revealed, for constant load exercise over time (12 and 15 repetitions) from an initial CR10 rating of 4 to 6, a linear increase of 1 scale point for every 3 to 4 repetitions. In conclusion, Borg's equation has been used to set theoretical estimates of a %1RM. Relevant to current practice guidelines was the ability to set appropriate loads in relation to performing recommended numbers of repetitions (e. g., if the CR10 rating is >6 after 2 repetitions, the weight is likely be too heavy to complete 12 to 15 repetitions).

Show all publications by Gunnar Borg at Stockholm University

Last updated: November 30, 2018

Bookmark and share Tell a friend