Differences in the course of physiological functions and in subjective evaluations in connection with listening to the sound of a chainsaw, and to the sounds of a forest
The data relate to a manuscript that will be published in Frontiers in Psychology. Frontiers Media S.A. doi: 10.3389/fpsyg.2022.775173 Names of variables: Sound: 1 = forest; 2 = chainsaw ID: the numerical identification of examined person (EP) AbR: abdominal respiration ThR: thoracic respiration AbRR or ThRR - respiratory rate AbVTR or ThVTR: relative tidal volume change SCL: skin conductance level FT: finger skin temperature BVP: blood volume pulse BVPA: blood volume pulse amplitude HR: heart rate RRit: R-R interval trends Alpha: peak alpha frequency HRV Peak HF: peak high-frequency HRV ASD: absolute values of standard deviations. Example AbRASD: absolute values of standard deviations of abdominal respiration HRV_nosignif - other not significant heart rate variability parameters; from ibi...to rLFHF_B We defined 4 time sections for the statistical comparison of differences among the average values of sounds. Periods, defined in MANOVA as a covariate, were not fixed but determined according to literary sources and similar findings from the line charts. Period1: a control period, 10 s before turning on the sounds. Period2: from 0 to 79 s Period3: from 80 s to 209 s (up to 3 ½ min), lasting 130 s (ca. 2 minutes). Period4: from 210 s to 390 s (up to 6 ½ min), lasting 180 s (3 minutes). (Sounds) Evaluation: AS - activation scale VAS - visual analogue scale In the AS method, one of ten polarity statements are selected on a scale from 1 to 11. In the VAS method, the respondent ticks his/her current mental and physical state “intuitively” on the line (encodable on a scale 0 – 200 mm). ASpre, VASpre - prior to the exposure of acoustic stimuli ASpost, VASpost - after the exposure of acoustic stimuli
Steps to reproduce
Data was evaluated by using statistical-analytical computer programs SPSS (version 25), Statistica (version 13.2) and Matlab (version 2015b). We made descriptive statistics, normality plots with data normality tests: Shapiro-Wilk W, Wilcoxon test and Kolmogorov-Smirnov with Lillefors correction. Since the data does not always exhibit normal distribution, a comprehensive comparison was made of medians, means and retransformed means by means of Box-Cox transformation. We also considered the range of quartiles, with data distribution evaluated based on boxplots, P-P plots, Q-Q plots and histograms. Nevertheless, results following out of parametric and nonparametric statistical methods were very similar. We also used Levene's Test for Equality of Variances and T Tests (Independent-Samples T Test) and for the control their nonparametric analog Mann-Whitney U for two-independent-samples. Kolmogorov-Smirnov Z (verification of Mann-Whitney U) was conducted only in sporadic cases of nonparametric data distribution. In order to compare relations between the physiological functions and the sounds and periods, we used the method of Repeated measures first. Nonetheless, after significant findings from Mauchly’s Tests of Sphericity (P < 0.001), we proceeded to the application of MANOVA. For the separate analysis of HRV, the research team used the raw data for additional calculation of R-R intervals. The R-R intervals of the BVP curve are expressed in the range of 30 – 200 bpm as well, but with a higher resolution accuracy of 0.004 bpm. Impaired records from the BVP changes were excluded from the analysis. Peak detection in records was performed using a custom-made algorithm based on the shape analysis of filtered records. Finite impulse response low-pass filter with cut-off frequency of 4 Hz and 30dB ripple performed by the Chebyshev window was used to suppress noise and artifacts. Only those peaks that achieved a height of at least 10% of average peak value and a spacing higher than 375ms were evaluated as valid peaks of pulse waves. Peak detection results were manually reviewed by the experimenter. The series of R-R intervals were derived from peak positions. HRV was analysed on the R-R interval series (Barbieri et al., 2005). Pre-processing of R-R series included the removal of ectopic intervals and R-R series detrending. Ectopic intervals differing by more than 20% were removed from the R-R series. The series of R-R intervals were detrended by wavelet packet decomposition. HRV was analysed in time, geometric and frequency domain and in non-linear domain. The heart rate variability analysis software (HRVAS) plug-in for Matlab software was used to compute HRV parameters (Task Force, 1996). Standards for the evaluation of HRV unify the length of the assessed tachograph section to 5 minutes so that both short- and long-term HR changes can be optimally captured for the correct quantification of which a time section of at least 2 minutes is needed.