Examining age differences in relations between physiological and self-reported stress data
Running Head: THE RELATIONSHIP BETWEEN SELF-REPORTED STRESS 1
THE RELATIONSHIP BETWEEN SELF-REPORTED STRESS MEASURES AND PHYSIOLOGICAL STRESS MEASURES; EXAMINING IF AGE FACTORS INTO THE RELATIONSHIP
Presented to the Campuswide Honors Program
University of California, Irvine
In Partial Fulfillment
of the Requirements for the
Campuswide Honors Program
TABLE OF CONTENTS
SELF-REPORTED AND PHYSIOLOGICAL DATA..........................................9
AGE AND STRESS.................................................................................10
DETAILS ON MEASUREMENTS...............................................................16
REFERENCES CITED .....................................................................................27
Firstly, I would like to thank my thesis advisor, Professor Reich, for agreeing to help oversee this thesis and work with me on it.
Additionally, I would like to thank my family and friends for their continuous help and support as I navigated this paper.
Stress is an aspect of life most people encounter, and there can be many causes for it. However, not everybody consciously evaluates their stress levels on a constant basis. This thesis attempts to answer the questions: 1. Is there a correlation between self-reported (subjective) stress measures and heart rate variability (HRV) as a physiological stress measure, and 2. Is age a factor in the relation between youths' subjective stress measures and their physiological stress measures? Participants were n=129 youths between the ages of 10 and 16 who were tested as part of a TSST experiment about teens and technology. The data analysis involved testing for correlations between HRV and self-reported stress rankings and testing for correlations between HRV and self-reported stress rankings when controlling for age, grade, or birth year as variables. Independent t tests showed gender as a possible confound for self-reported data- females tended to self-report higher stress levels than males did. The results of this experiment aligned with previous research, suggesting that the youths' subjective data had significant relations with their physiological data. Age was also found to have a significant relation when testing for the correlation between self-reported and physiological data. This research contributes to the literature about subjective and physiological data by exploring the relations between the two in a lesser-studied population- youths. Future research can focus on investigating if different stage of puberty between the genders also factor into the correlation between youths' subjective, self-reported stress levels and physiological data.
Historically, physiological and self-reported data have been used to measure biological aspects of life. For example, people wear fitness trackers to record their heart rate or daily activity, and there exist various apps that allow users to log their food intake- whether by naming the food type, scanning a barcode or taking a picture of the meal. Each of these healthy habits can be empirically validated, since data logged in at one application should match, or closely match, data from another location (Teixeira, Voci, Mendes-Netto, & da Silva, 2018). Stress is another biological aspect of life, but is a much more subjective process than logging food, since it can include aspects such as physiological responses, emotional reactions, and/or mental interpretations (Campbell & Elhert, 2011). Since different people have different biometrics, this means the same physiological stimulant can lead to a different mental interpretation from one person to the next. This, in turn, can lead to a different physical manifestation and reaction to the stressful situation (Semmer, Zapf, & Dunckel, 1995). All these different factors lead to a subjectiveness of self-reported data, which makes it more difficult to analyze self-reported data for stress, compared to self-report of something like one's caloric intake.
As an alternative, physiological methods, such as wearing a heart rate monitor to determine heart rate variability (HRV) or testing cortisol levels in saliva, can also be used to measure stress, but the collecting and analyzing of physiological data can also be more invasive or expensive than collecting and analyzing self-reported data (Makivi?, Niki?, & Willis 2013; Sharpley 1998) While both physiological and self-report methods of measurement can be, and are used, and the relationship between HRV and self-reported data have been examined more often than that of cortisol and self-reported data, it is still unclear whether or not the relationship between physiological stress data and subjective stress data translates across all age groups- that is, from youths to adults.
Prior research has used different methods of measuring self-reported stress. For the purposes of this paper, "self-reported" includes any type of data in which answers to questions would be considered subjective. Examples of these include: the State Trait Anxiety Inventory (STAI) survey (Hellhammer & Schubert, 2012; Roemer, Borkovec, Posa & Borkovec, 1995; Leininger & Steel, 2012; Julian 2011; Spielberger, Gorusch, & Lushene 1961), Daily Stress Inventory (Brantley, Waggoner, Jones, & Rappaport, 1987; Brantley, Dietz, McKnight and Jones, & Tulley, 1988), or Perceived Stress Reactivity Scale (PSRS) (Schlotz, Yim, Zoccola, Jansen & Schulz, 2011; Morgan, Umberson, Hertzog, 2014 ). While this is not a comprehensive list of all self-report methods, it is a sample of what has been used in the past. The experiment for this paper used a 7-point survey (rating from 0-6) as a subjective method for participants to self-report their stress levels. Most studies have found these different self-report measures to be internally valid; moreover, different studies of these self-report methods have found the results of most self-report measures to be significantly correlating with physiological data, such as endocrine or cortisol levels (Barnes, Harp, & Sik Jung, 2002).
Physiological data have often been used as a more objective measure of stress. Physiological data comes from biological aspects of a participant and typically utilize some form of scientific or health-related instruments for analyzing or collecting, such as heart rate monitors. It is therefore seen as more reliable, since the instruments or devices used to collect the data can automatically measure and analyze the information in increments that would otherwise be tedious for humans (Ding, Hotho, Holmberg, Fuss, & Sperlich 2016). For the purposes of this paper, "physiological data" is defined as biological data that is collected using health-related equipment. Prior studies have used endocrine levels in urine (Brantley, Dietz, Mcknight and Jones, & Tulley, 1988; Cox, 1985) or blood (Vaernes, Ursin, Darragh, & Lambe 1982), cortisol levels in saliva (Leininger & Skeel, 2012; Burke, Davis, Otte, & Mohr, 2005), or variability in heartrate (Taelman, Vandeput, Spaepen, & Van Huffel, 2009) to measure stress. Cortisol levels in saliva and heart rate variability have both been often-utilized types of physiological data used for measuring stress, and similar to the experiment done by Taelman et. al, the experiment discussed in this paper utilized heart rate variability as the physiological aspect of stress.
Heart rate variability, also known as HRV, is the variation in time in between heartbeats. It can be used to examine stress, physical fitness, or potentiality for heart diseases, and is found by calculating the root mean square of the successive differences (RMSSD) between neighboring RR intervals (Buccelletti, Gilardi, Scaini, Galiuto, Persiani, Biondi, Basile, & Gentiloni Silveri, 2009). More simply put, it calculates the difference in time between heart beats in heart rate data. For the purposes of this experiment, HRV was used to examine stress. This paper aims to examine the potential correlation between heart rate variability and a participant's self-report data.
Self-reported and physiological data
Previous research (Hellhammer & Schubert, 2012; Polheber & Matchock, 2013; Campbell & Elhert, 2011) has found mixed results regarding whether or not self-reported levels of stress consistently reflect physiological measures of stress, such as heart rate (HR) or cortisol levels in saliva. For instance, using the Trier Social Stress Test (TSST) to induce stress, Polheber et al. (2013) measured stress with both self-reported and physiological, objective measures. The State Trait Anxiety Inventory (STAI) questionnaire was used as a self-reported stress, while monitoring heart rate (HR) and testing cortisol levels were used for objective measures of stress. The authors found that, "STAI State or Trait anxiety did not significantly correlate with any measures of cortisol or HR" (p. 863), and "STAI measures were not associated with any measure of cortisol production" (p. 865). Similarly, when Campbell and Elhert (2011) set out to examine the associations between physiological and emotional (self-report) responses to the TSST, of the 49 studies they examined, only 25% of them (approximately 12) had "significant relations" between the physiological measure (cortisol) and a subjective, self-report measure of perceived emotional stress. In fact, in over 50% of the TSST studies Campbell and Elhert examined, (approximately 26) there was either "no predictive power" or "no correlation" between the subjective self-report data and physiological data. On the other hand, Hellhammer and Schubert (2012) found that while using a TSST to induce stress, their participants' self-report data during their TSST predicted cortisol levels for perceived stress. However, in Hellhammer and Schubert's experiment, the authors' method of self-report data was on a much larger scale- participants ranked their perception of stress, anxiety, and insecurity on a line marked from 1 to 100. Such a wide range could potentially reduce room for error and make it easier for participants to distinguish their stress level, as opposed to on a 7-point scale, which is a much a smaller range. Furthermore, their experiment was with adults, and it would be interesting to see if the results they found would also translate to a younger population. Another significant finding, though, from Yim, Quas, Rush, Granger, and Skoluda (2015) was that, in their trial with adults, there were significant associations between self-reported stress and the a-amylase levels in saliva. However, since Yim and colleagues' trial, like most others, was with adults, it still leaves an age gap in the research- one for individuals in their pre-teen through early teenage years.
One interesting result was the analyses from the experiment that Luettgau, Schlagenhauf, and Sjoerds (2018) carried out, which showed a "no significant relationship" (p. 31) between the subjective stress measure and salivary cortisol levels. Similar to Yim and colleagues' trial, participants ranked their stress on a Visual Analogue Scale (VAS), on a scale from 1-100.
Age and Stress
Most of these previous studies have focused on older adolescents or adults, which leaves a gap in the research. This study, done on participants aged 10-16, covers three age ranges, as defined by Amirkhan and Auyeung (2007): Pre-teens (9-12), early teens (13-15), and late teens (16-18). The first age range, pre-teens, has not been studied as much, and therefore, this study is unique, as the ages of the participants in the study encapsulate parts of age ranges that are both lesser studied (10~12), and more studied (16), and an age range of individuals who have been more studied (13~15). There is no shortage of research on the validity of self-report data; however, when specifically investigating the validity of pre-teens and self-report data, there is a noticeable difference in the amount of research when compared to that of teens and/or adolescents. This paper fills the gap, since it investigates how much the participants' self-report data correlates to their heartrate data as age progresses.
This study is different from previous work, which have focused on higher age ranges. It would be worth finding out if there is a difference in how age affects stress perception amongst teens and pre-teens. Additionally, in prior experiments using heart rate variability, participants were generally older. It is therefore worth investigating further whether these differences in measurement methods and age have an effect on an individual's stress levels, and/or people's perception of their own stress.
This study aims to answer two questions: Is there a significant relation between HRV (a measurable physiologic feature) and participants' self-reported stress data, and does age have an effect on how self-report data correlates to physiological data? The prediction is that there will be a weak correlation between participants' physiological and self-reported stress data, but no correlation between the age of participants and the correlation of their self-reported and physiological stress data.
Participants were students in the Orange County and Los Angeles County area between the ages of 10 and 16. Private and public schools were contacted in both counties in regards to potential interest and participation, but most participants were recruited from afterschool centers and activities. Other forms of recruitment included flyers, word of mouth, and snowballing. Consent forms were left with the directors of various programs, who then helped distribute the forms to interested students, who brought them home for parents to sign. If the student returned the form, they were allowed to participate in the experiment. All participants were from a TSST experiment about teens and technology and were told they could win a pair of movie tickets if they participated.
In total, 129 participants between the ages of 10 and 16 (grades 6-9) were tested. There were 63 male participants and 66 female participants, all of whom came in to the experiment in same-gendered pairs. The average age of the participants was 12.4 years old, with a standard deviation of 1.24. Participants who were tested identified as various identities, including African American, East Asian, East Indian, Hispanic, Middle Eastern, Native American, Pacific Islander, Southeast Asian and White. Some participants (25 total) considered themselves to be of more than one ethnicity.
The inclusion criteria for the experiment was: 1. participants had to join the experiment with a friend of the same gender who was not a sibling, 2. participants had to be between the ages of 10-16, and 3. participants had to understand English well enough to understand directions and complete the tasks. If any participants were excluded, it was because they chose to not continue with the experiment.
In order to assess the relationship between self-report and physiological measures of stress, the Trier Social Stress Task (TSST) method of experimentation was used to put participants into a stressful situation. This TSST experiment data came from a larger experiment about teens and technology. The TSST experiment took place in three parts: the baseline (T1), the TSST task/stressor portion (T2), and the post-stress task (T3). Experimenters recorded the time after each portion of the experiment so each section of the experiment had a distinct timeframe to reference when analyzing HRV data.
Upon arrival, pairs of participants were greeted, separated into different rooms, and then talked through an assent form with an experimenter. Once the assent form was signed, the participants were asked to silence or turn off any devices and leave their belongings in a corner. The experimenter then gave the participant a Polar H10 Heart Rate Sensor on a chest strap for detecting heart rate and a Polar V800 Watch for recording the heart rate sensor's data, and then took the participant to a nearby bathroom so they could put on the equipment. After the participant finished putting on the equipment, they were led back to the room to begin the baseline portion of the TSST experiment.
During the baseline portion, participants watched a 5-minute, relaxing video of nature. They were told to simply relax and enjoy the video. When the video ended, they were given a T1 form to rate their mood level from -3 to 3 and their stress level from 0 to 6. The experimenter then left the room and switched places with the experimenter assisting the other participant, to begin the stressor portion of the experiment.
During the stressor portion of the experiment, the experimenters come in and inform participants they must give a 5-minute speech on why they believed the should win movie tickets. Participants were given 3 minutes to write notes, but the paper was taken away during the actual speech and participants were not allowed to use the notes written on it. During the 3 minutes, experimenters leave the room to retrieve a video camera. After the three minutes are over, experimenters inform the participants that their speech would be "video recorded," as it was a competition- the "footage" would be reviewed, and whoever had the best speech performance would win the pair of movie tickets. After the speech, participants were given a math task, counting backwards by three from 1022. After 5 minutes of the math task, participants were given the T2 form, which once more had them rate their mood level from -3 to 3 and their stress level from 0 to 6.
The post-stress task had the participants text the friend they came with on Google Hangouts for 5 minutes, watch a 5-minute video on brick-laying, or do nothing for 5 minutes. After the 5 minutes were over, participants were given a T3 form to rate their mood level from-3 to 3 and their stress level from 0 to 6 a final time. Afterwards, the experimenter informed them of the experiment's end, and they were once more taken to the bathroom, to take off the heart rate equipment. Participants were then debriefed, given a pair of movie tickets as a show of gratitude, and allowed to retrieve their belongings.
Demographics. Participants' demographics were collected through an online survey, which was created by the experimenters. Variables collected that were relevant to the study included age, birth year, gender, and grade level.
Self-report measures. After each section of the experiment (baseline, TSST, post-TSST) participants completed a seven-point survey (T1, T2, or T3) that asked them to rank their mood on a scale from -3 to 3, and their stress from 0 to 6. The responses to the questionnaires were the subjective stress measure, since each participant would have their own interpretation of how they felt. The experimenters coded and recorded the participants' responses on an online spreadsheet after the experiment.
Physiological measures. This experiment used heart rate variability as a physiological measure of stress. Heart rate was monitored using Polar H10 Heart Rate Monitors, and data was recorded on Polar V800 watches. Polar V800 watches can connect to a computer via USB, and the data recorded on the watches is uploaded onto Polar's online system, Polar Flow. Polar Flow can automatically calculate heart rate variability by computing the root mean square of successive differences (RMSSD) for whatever time frame the user highlights to choose. RMSSD in this case was the same as HRV, therefore, that is how the HRV was calculated in this experiment. All values were then manually recorded into an online Excel spreadsheet by the experimenters.
Since the technology did not always work, there were some participants for whom there was missing or unanalyzable data. For that reason, if the data was missing or unanalyzable, it was excluded. Unanalyzable data would include heart rates of 0, or irregular spikes in the heart rate data that would have skewed the RMSSD or HRV calculation process.
Details on the measurement:
On the T1, T2, and T3 questionnaires, participants were asked to rank their mood (-3-3) and stress levels (0-6) after each stage of the experiment. A rank closer to -3 meant a more negative mood, and a rank closer to 0 meant less stress.
RMSSD measured heart rate variability (HRV). HRV is controlled by the body's autonomic nervous system (ANS), which is divided into two parts- the sympathetic and parasympathetic, also known as the "fight or flight," or "relaxation" responses (Buccelletti, Gilardi, Scaini, Galiuto., Persiani, Biondi, Basile & Gentiloni Silveri, 2009). When a person is in fight or flight mode, their HRV is much lower, which shows a higher level of stress. Since HRV measures the time difference between heartbeats, a low variation means the body is slower to react to new situations, which, therefore, denotes higher stress. A higher HRV shows more resilience towards stress and situations, as it shows the person is able to quickly adapt, and therefore suggests lower stress levels (Campos 2017).
Since the experiment used Polar heart rate monitors and Polar watches, the heart rate data collected through the equipment was automatically backed up onto Polar's website, Polar Flow. Upon accessing the data, the Polar Flor system shows an individual's heart rate throughout the experiment. The system also measures heart rate variability for any given time frame throughout the experiment, by calculating the value of the root mean square of the successive differences (RMSSD) for any portion of the data that a user chooses to highlight. Therefore, the experimenters used the Polar Flow system to calculate the RMSSD by highlighting the time frames for each part of the experiment, and then manually recording the numbers into an online Google spreadsheet. A high value for RMSSD meant the individual had a lower stress level, whereas a low RMSSD value meant the individual had a higher stress level, since RMSSD is the process of finding HRV.
In order to test the first hypothesis, Pearson's correlation analysis tests were used to determine if participants' self-report had a significant correlation to their physiological, HRV data. The test was run three times- once for each time frame. Each time, variables being compared were the self-report and the HRV data for each time frame- S1, S2, and S3, as well as the RMSSD (HRV) data for the baseline, TSST, and post-TSST. Two independent samples t test were also done to examine if gender was a confound for the HRV or self-report variables.
To test the second hypothesis, a Partial correlations test was used, with age as a constant, to determine if age was a significant factor in how the participants self-reported their stress. Additionally, since age is typically linked to birth year and grade level, those two variables were also tested in separate partial correlations tests to determine if age played a role in how participants' self-reported stress related to their physiological stress data. In each of these tests, the dependent variables were the self-report scores and physiological data, while birth year, grade level, or age were set as control variables.
All data analysis for this paper was done using IBM's SPSS version 25.
The demographic information for the sample of participants in this experiment is shown in Table 1. Information for one male was excluded because upon debriefing, he decided not to continue with the experiment. Since all participants who went through the experiment did so in dyads, his withdrawal from the experiment resulted in an odd number of males.
In order to test hypothesis 1, which suggested there would be a weak correlation between self-report and the physiological HRV data, three Pearson's correlations tests were run. The first one tested self-report and HRV at for the first time interval, the baseline measure. The second test examined correlation between self-report and HRV at the second time interval- the stressful TSST task, and the final test examined correlation between self-report and HRV for the post-stress task. For the baseline measure, there was no significant correlation (? = .096) found between the participants' self-report and physiological data. For the second time frame, there was also no significant correlation (? = .052) between the participants' self-report and physiological data. Finally, in the last time frame (post TSST) there was also no significant correlation (? = .480) between the self-report and HRV data. These results suggest that the first hypothesis was incorrect, since there were no significant correlations between the participants' self-reported stress level and physiological stress level across the three time frames.
Hypothesis 2 predicted that there would be no significant correlation between participants' age and the correlation of self-report and physiological HRV data. In order to test this, a Partial Correlations test was run, using the self-report and physiological data for each time frame as variables while using age as a controlled variable. In the first test, when age was set as a controlled variable, there was no significant correlation (? = .117, r =.145) between the participants' self-report and physiological data. For the second test, there was a slight negative correlation (? = .033, r = -.197) between the two sets of data. Finally, in the third test, there was no significant correlation (? = .422, r = -.074) between the two sets of data. The results from the first and the last test showed that the hypothesis was partially correct. When age was controlled as a variable during the baseline and post-stressor task, there were no significant correlations between self-report and physiological data. However, during the stressful task, there was a weak negative correlation between the two types of stress data.
Since age is related to birth year and grade level, tests were also run to examine if birth year or grade were also a potential factor regarding the correlation between self-report data and physiological, HRV data. Similar to the process used to test for correlation with age, three partial correlations tests were run, one for each timeframe, with birth year as a control variable and testing for correlation between HRV and self-report data. In the first test, there was no significant correlation (? = .095, r = .155) between the two sets of data. In the second test, there was no significant correlation (? = .085, r = -.160) between the physiological and self-report data, and in the final test, there was also no significant correlation (? = .516, r = -.060).
For grade level, three partial correlations tests were once again run. For each of the tests, the control was set to grade level and the variables were set to self-report scores and HRV. In the first test, there was no significant correlation (? = .139, r = .138) between the self-report and HRV. The second test also had no significant correlation (? = .083, r = -.161), as was the case with the third test (? = .439, r = -.072).
Two independent t-tests were run to test for confounding variables. The first test examined if gender was a potential confound for HRV. The independent variable was gender, while the dependent variable was HRV. The results showed no significant relation, suggesting that gender was not a confounding factor of HRV. The second test examined if gender was a potential confound for the subjective, self-report data. The dependent variable was self-report data, and the independent variable was gender. The results found that there was a significant relation between gender and the self-report scores at times 1 and 2 (baseline and TSST). The results suggest that gender may have been a confounding variable in how participants reported their own stress, and therefore, the relation between the self-reported data and HRV data. The descriptive statistics of how each gender self-reported their stress are shown in Table 2.
The purpose of this experiment was to determine: 1. if self-reported stress data in general correlated to physiological stress data, and 2. if age had a factor in how well youths self-reported their stress levels, when those self-reported data are compared to physiological data. This was done in order to contribute to the literature regarding age in correlations between self-report and physiological stress data. The initial hypotheses stated that 1. there would be a weak correlation between self-reported and physiological data for stress, and 2. there would be no significant correlations between age and the correlations of self-report and physiological data.
For the first hypothesis, it was determined that there was no significant correlation between the participants' self-reported and physiological data for stress at all three times. However, it is worth noting that the experiment worked with a relatively small sample, and at the time of stress (self-report immediately after the TSST, and HRV during the TSST), there was a .052 significance value and -.178 correlation value between the HRV and self-report data. Had the sample been slightly larger, this value may have become significant. Additionally, due to technical difficulties, not all participants' data was included in the analyses. Therefore, it may be that if the experiment had a larger sample, there would have been a statistically significant negative correlation between the two values. Given these results, this leads to a new theory- that it may be easier for a person to assess their stress when they are put under a stressful situation.
For the second hypothesis, there was a significant, negative correlation between the self-report and physiological data variables at the second time of self-report when age was controlled as a variable. The .033 significance value and -.197 correlation value show a weak, negative correlation between the self-report and physiological data. However, there were no significant correlations at the baseline or post-TSST task.
While grade level and birth year did not have significant correlations where self-report and physiological data were concerned, it is still important to note that the significance value for both variables were closest to ?=.05 at the second time interval- right after the TSST task. This is noteworthy because previous research has found that subjective measures of stress are closest related to physiological measures during the TSST, but neither before nor after (Hellhammer & Schubert, 2011). Therefore, these result across the three variables (age, grade, and birth year) are in line with previous research. Even though there were weak, negative correlations at the time of stress, there were no significant correlations between the self-report data and physiological data at the other times. As the title of the research Hellhammer and Schubert (2011) conducted states, "The physiological response to Trier Social Stress Test relates to subjective measures of stress during but not before or after the test" (119).
The slight negative correlation from the TSST portion of the experiment in each of the tests is expected, because as stress increases, the HRV is meant to decrease. A lower HRV shows the body's inability to properly adapt to new stimuli, which causes stress. On the contrary, a higher HRV shows that the body is able to adapt to new stimuli, which, consequently, does not cause stress (Campos 2017).
The t-test findings that showed gender as a potential confound for the self-report scores is in line with the findings that Kelly, Terka, Anderson, Price, and Carpenter (2007) reported, which was that women tend to self-report higher levels of stress, distress, or anxiety than when compared to men. The mean self-report scores from females were higher than those of the males by approximately half a point or more, which is a rather large difference on a 7-point scale (from 0 to 6). Armed with this knowledge, it is plausible that gender could very well have been a confound for the whole experiment, since it was a confound for the self-reported, subjective stress data.
On the other hand, the t test results that showed gender to be not a confound to HRV during the stressor portion of the TSST (?=.886) is an interesting conclusion, since the results from Kudielka, Kirschbaum, Hellhammer, and Kirschbaum (2004) reported that females tended to have higher heart rates during their stress tasks than men. While heart rate is not the same as heart rate variability, the two are still related, as you cannot calculate heart rate variability without first having the heartrate. However, it is possible that nonsignificant results from the t test in this experiment may have been because, due to the occasional technical difficulties, there were not enough participants with data to find significance (n = 58 females, and n= 62 males). It is worthy to note the difference in results here, because having a difference of n=4 females in a sample size so small could significantly affect the data.
This experiment has a few major limitations to it, the first of which being sample size. With only 130 total participants and one not wishing to continue after being debriefed, using the data of 129 youths is not a particularly representative sample. Recruitment was limited to nearby schools, afterschool activities, and word of mouth, since recruiting through schools required district approval before experimenting. Therefore, it was difficult to find qualifying students who were interested in participating. A larger sample of youths could potentially yield stronger or more significant results.
Additionally, the age range experimented on here is the prime range for youth's puberty. Even though pubertal status was not part of the data collected in this experiment, it is possible that puberty could relate to how youths subjectively assess their stress levels, because puberty can be related to emotional maturity. Therefore, depending on the individual's emotional maturity, it could also affect how they subjectively analyze their feelings. Further research could benefit from investigating how puberty or emotional maturity would relate to subjective or physiological stress measures.
Another limitation was technological difficulties. Since the physical stature of people aged 10-16 can vary, the heart rate strap did not always work optimally. Therefore, there were times for some people (n=9) at which the heart rate data was off or not collected correctly. This included spikes in heart rate or a heart rate of 0, which led to exclusion of the data. The exclusion also reduces the amount of heart rate data available- with the smaller sample, it may have been even more difficult to find significant correlations.
Despite finding non-significant results, this study was useful and contributed to the literature regarding subjective and physiological manifestations of stress, because it covered an age range that is currently not as thoroughly studied. It also highlighted that gender is a potential confound in how people subjectively self-report their stress. It may be useful to further study how pubertal status and gender are related to subjective and physiological measures of stress, especially using noninvasive methods such as heart rate variability.
Amirkhan, J., & Auyeung, B. (2007). Coping with stress across the lifespan: Absolute vs. relative changes in strategy. Journal of Applied Developmental Psychology, 28, 298-317. doi:10.1016/j.appdev.2007.04.002
Barnes, L., Harp, D., & Jung, W. (2002). Reliability generalization of scores on the Spielberger State-Trait Anxiety Inventory. Educational and Psychological Measurement, 62, 603-618. doi:10.1177/0013164402062004005
Brantley, P. J., Dietz, L. S., Mcknight, G. T., Jones, G. N., & Tulley, R. (1988). Convergence between the daily stress inventory and endocrine measures of stress. Journal of Consulting and Clinical Psychology, 56(4), 549-551. doi:10.1037/0022-006X.56.4.549
Brantley, P. J., Waggoner, C. D., Jones, G. M., & Rappaport, N. B. (1985, October). A daily stress inventory: Development, reliability, and validity. Journal of Behavioral Medicine, 10(1), 60-74.
Burke, H. M., Davis, M. C., Otte, C., & Mohr, D. C. (2005). Depression and cortisol responses to psychologicalstress: A meta-analysis. Psychoneuroendocrinology, 30, 846-856. doi:10.1016/j.psyneuen.2005.02.010
Gilardi, E., Scaini, E., Galiuto, L., Persiani, R., Biondi, A.,
Campbell, J., & Ehlert, U. (2012). Acute psychosocial stress: Does the emotional stress response correspond with physiological responses? Psychoneuroendocrinology, 37, 1111-1134. doi:10.1016/j.psyneuen.2011.12.010
Campo, M. (2017, November 22). Heart rate variability: A new way to track well-being. In Harvard Health Publishing. Retrieved June 15, 2019, from https://www.health.harvard.edu/blog/heart-rate-variability-new-way-track-well-2017112212789
Cox, T. (1985). The nature and measurement of stress. Ergonomics, 28(8), 1155-1163. doi:10.1080/00140138508963238
Ding, P., Fuss, F. K., Holmberg, H., & Sperlich, B. (2016). Recommendations for assessment of the reliability, sensitivity, and validity of data provided by wearable sensors designed for monitoring physical activity. JMIR MHealth and UHealth, 6(4). doi:10.2196/mhealth.9341
Hellhammer, J., & Schubert, M. (2011). The physiological response to Trier Social Stress Test relates to subjective measures of stress during but not before or after the test. Psychoneuroendocrinology, 119-124. doi:10.1016/j.psyneuen.2011.05.012
Julian, L. J. (2011). Measures of anxiety. Arthritis Care & Research, 63(S11), S467-S472. doi:10.1002/acr.20561
Kelly, M. M., Tyrka, A. R., Anderson, G. M., Price, L. H., & Carpenter, L. L. (2008). Sex differences in emotional and physiological responses to the Trier Social Stress Test. Journal of Behavior Therapy and Experimental Psychiatry, 39, 87-98. doi:10.1016/j.jbtep.2007.02.003
Kudielka, B., Buske-Kirschbaum, A., Hellhammer, D. H., & Kirschbaum, C. (2004). Differential heart rate reactivity and recovery after psychosocial stress (TSST) in healthy children, younger adults, and elderly adults: The impact of age and gender. International Journal of Behavioral Medicine, 11(2), 116-121. doi:10.1207/s15327558ijbm1102_8
Leininger, S., & Skeel, R. (2012). Cortisol and self-report Measures of Anxiety as Predictors of Neuropsychological Performance. Archives of Clinical Neuropsychology, 27, 318-328. doi:10.1093/arclin/acs035
Luettgau, L., Schlagenhauf, F., & Sjoerds, Z. (2018). Acute and past subjective stress influence working memory and related neural substrates. Psychoneuroendocrinology, 96, 25-34. doi:https://doi.org/10.1016/j.psyneuen.2018.05.036
Makivi?, B., Niki?, M. D., & Willis, M. S. (2016, June). Heart rate variability (HRV) as a tool for diagnostic and monitoring performance in sport and physical activities. Journal of Exercise Physiology, 16(3), 103-131.
Polheber, J. P., & Matchock, R. L. (2014). The presence of a dog attenuates cortisol and heart rate in the Trier Social Stress Test compared to human friends. Journal of Behavioral Medicine, 860-867. doi:10.1007/s10865-013-9546-1
Roemer, L., Borkovec, M., Posa, S., & Borkovec, T. D. (1995). A self-report diagnostic measure of generalized anxiety disorder. Journal of Behavior Therapy and Experimental Psychiatry, 26(4), 345-350.
Schlotz, W., Yim, I. S., Zoccola, P. M., Jansen, L., & Schulz, P. (2011). The perceived stress reactivity scale: Measurement invariance, stability, and validity in three countries. Psychological Assessment, 23(1), 80-94. doi:10.1037/a0021148
Semmer, N., Zapf, D., & Dunckel, H. (1995). Assessing stress at work: A framework and an instrument. Work and Health: Scientific Basis of Progress in the Working Environment, 105-113.
Sharpley, C. E. (1998). Psychosocial stress-induced heart rate reactivity and atherogenesis: Cause or correlation? Journal of Behavioral Medicine, 21(5).
Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). STAI manual for the State-Trait Inventory. Palo Alto.
Taelman, J., Vandeput, S., Spaepen, A., & Van Huffel, S. (2009). Influence of mental stress on heart rate and heart rate variability. In J. Vander Sloten, P. Verdonck, M. Nyssen, & J. Haueisen (Eds.), 4th European Conference of the International Federation for Medical and Biological Engineering (Vol. 22, pp. 1366-1369). Berlin, Germany: Springer, Berlin, Heidelberg. doi: 10.1007/978-3-540-89208-3_324
Teixeira, V., Voci, S. M., Mendes-Netto, R. S., & da Silva, D. G. (2018). The relative validity of a food record using the smartphone application MyFitnessPal. Nutrition & Dietetics, 75, 219-225. doi:10.1111/1747-0080.12401
Vaernes, R., Ursin, H., Darragh, A., & Lambe, R. (1982). Endocrine response patterns and psychological correlates. Journal of Psychosomatic Research, 26(2), 123-131.
Yim, I. S., Quas, J. A., Rush, E. B., Granger, D. A., & Skoluda, N. (2015). Experimental manipulation of the Trier Social Stress Test-Modified (TSST-M) to vary arousal across development. Psychoneuroendocrinology, 57, 61-71. doi:10.1016/j.psyneuen.2015.03.021