Stress Levels of City College Students
Introduction (Daniel)
Stress is defined as any feeling of emotional or physical pain that causes tension, it can derive from many factors that induce feelings of frustration, anxiety, anger, and nervousness. MedlinePlus (2019), classifies stress into two categories: acute and chronic stress, acute stress is short term stress that goes away quickly and induces the fight or flight response that is typically beneficial in helping avoid potentially dangerous situations such as having to swerve out of the way when a child runs into traffic. Chronic stress is classified as ongoing stress that is often left unmanaged, although stress is not tangible the effects of stress are substantial and could lead to health problems such as high blood pressure, obesity, depression and skin problems (MedlinePlus, 2019).
Students especially fall victim to chronic stress trying to maintain an impressive academic record, juggle social relationships, handle finances and maintain familial relationships (Brougham et al., 2009). There is no single underlying factor influencing stress levels in students rather multiple factors multiplying the stress students experience. In our experiment, we investigated how CCNY college student’s gender, religious status, hours working, and credits attempting correlated to the stress level indicated by each volunteer. This experiment is particularly useful for college students because it helps identify and understanding exactly what factors are common among stressed college students potentially leading to help develop programs that are specialized in helping manage stress say for an amass of school work and part-time employment. Below are hypotheses made based on common assumptions.
a. There will be a significant difference in female and male stress levels.
b. Religious persons will have lower levels of stress compared to non-religious persons due to psychological outlets many religions provide (such as prayer).
c. Students working full time will have higher levels of stress compared to students working part-time.
Materials and Methods (Tenjing Sherpa)
A sample of 52 participants responses was collected for this experiment. We got all of our participants response online. We believed that the internet will keep the identity of participants private and anonymous, which will make them more likely to give accurate data. The participants ranged from age 17 to 23. It would have been better to have more participants, and more range, because we had 24 participants age 18 and, 11 participants age 19. So, the participants were mostly concentrated at the age group of 18 and 19, an even distribution of age would yield better outcomes for comparing the data with other categories.
We created an online survey using google survey (Appendix A), and we shared the link online on CCNY student Life mobile application. This is the app used by most CCNY students to ask questions and surveys. We also shared it on social media using CCNY Book market page on Facebook, which is similar to the CCNY app, but it has a very large group of CCNY student members. The sample size was 35 females, 14 males, and 3 preferred not to say. For our study, we decided to exclude the corresponding data of 3 of the participants responses, who did not prefer to identify their gender. According to APA guidelines, “gender refers to the attitudes, feelings, and behaviors that a given culture associates with a person’s biological sex”, and the gender can be male, female, or neutral (APA, 2012). For this reason, we had the gender as either male or female. But gender identity can refer to one’s sense of oneself as male, female, or something else (APA, 2011). Here the third variable can have multiple terms to describe oneself, which could also be non-binary. So, the participants can have many gender identities and, we cannot add all the possible gender to the list. Hence, we had the third option for gender as “prefer not to say”, which could be a neutral gender or any gender identity. Our purpose of the study is to see the impact of stress on gender, and for simplicity, we chose to study its effect on those who identify as either male or female. So, any answers besides male or female were regarded as the same and discarded for this study, which will allow us to have the desired result.
In the survey (Appendix A), the participants were asked questions about their age, gender, hours of work each week, credits taken this semester, and whether they considered themselves as religious or not. At the end of the survey, we asked the participants to rate themselves at how stressed they feel. We referred to stress as physical or emotional tension. The scale ranged from 1 to 5, with 1 being rarely stressed and 5 as extremely stressed. This scale represents the model of a Likert scale. The likert scale has a rating ranging for -5, -7, or -10 points scale. We could have used the 7-, or 10-point scales, but according to Dawes (2008), all three scales are equally desirable to obtain data for regression analysis. Since we were designing an online survey which could be done on a mobile device, having -5 point scales will be better than -7 or -10 point scales, because the option of having -5 point scales fit perfectly on the mobile device’s screen.
The survey we collected was meant for CCNY students, so this is a convenience sample because students are not randomly selected. Although the survey is answered by random students, the pool of participants we collected comes from the group who use CCNY student life mobile app, and CCNY book market Facebook page. The data we collected have high face validity because the title of the survey is “Stress level of CCNY students survey”, which makes the participants aware that the main purpose of the study is to examine the stress level and compare it to other factors which may contribute to it. Since we wanted to get a high number of participants, we made the survey short with seven questions. We obtain self-report data from the survey, but the participants could be biased when answering self-reported data. In our case, they may not be aware of their level of stress and, asking questions about their behavior to assess their level of stress would yield more accurate data. However, collecting behavior data requires more questions, and it would also have lower face-validity. Besides, the participants may not be willing to answer a survey if they are not aware of its purpose, so we chose to collect the self-reported data. I imported this data on MS-Excel because it allows the better function for data sorting and replacing the values of the data with a dummy variable for programming in SPSS. For example, I could replace all the value of Male as 0, and females as 1, and vice-versa. So, MS-Excel was used for data arrangement, and SPSS was used for analyzing and performing the test.
Procedure (Tenjing Sherpa)
All the participants who answered our survey question answered all the questions on the survey (Appendix A). We let the survey to remain active for about two weeks before we decided to stop collecting more data because we were not getting any more new data after that. Besides, we needed time to analyze and compute the data. After analyzing all the data, I found that we had no international students in our survey. So, we also had to discard that information, because we could not categorize the CCNY students into two groups of international students and non-international students.
We had usable data of 49 participants, which we programmed into SPSS and analyzed the data with the help of Lab T.A Anita Sicignano. The population sample did not have a normal distribution, has a small size (n<100), and since we were working with nominal data (e.g. gender) we used a non-parametric Test. We performed Independent samples Kruskal- Wallis test. This test does not assume a normal distribution, so the presence of an outlier does not have much of an effect on the result. This test assumes that the null hypothesis as having the median of all groups equal. The test determines whether or not there is a location shift in the distribution, and if the shift in distribution is significant. The alternative hypothesis of the test is not that one of the distributions has a different median. It is that one of the distributions has exactly the same shape as the others but is shifted up- or down-wards. So, calculating the median is part of the test but it is only used to test if the distribution of stress is similar across the categories of the other groups.
All of our analysis resulted in accepting the null hypothesis because the significance of it happening was more than 5%. So, we performed a post hoc test to see better differences between the groups.
Result
After performing the Kruskal-Wallis Test we got the following result:
- Gender and stress level
The significance level is .508, so we accept the null hypothesis that the distribution of stress level is the same across categories of male and female. This could have happened because the stress of work and college is perceived similarly by both the male and female. So, we rejected our hypothesis that there will be a significant difference in female and male stress levels.
- Being religious and stress level
The significance level is .789, so we accept the null hypothesis that the distribution of stress level is the same across categories of being religious or not. This could have happened if the participant did not pay attention to the portrayal of the character. So, we rejected our hypothesis that religious person will have lower levels of stress compared to non-religious persons due to psychological outlets many religions provide
- Being full time or part-time and stress level
The significance level is .940, so we accept the null hypothesis that the distribution of stress level is the same across categories of being full time and part-time students. This could have happened because we do not have a big sample size to see any significant difference among the groups.
- Age and stress level
The significance level is .270, so we accept the null hypothesis that the distribution of stress level is the same across categories of age. This could have happened because we do not have a big sample size to see any significant difference among the groups.
- Work and stress level
An independent sample, non-parametric test showed that the distribution of stress level is the same across all groups of working students (p>0.5).
So, we should reject our hypothesis that students working full time will have higher levels of stress compared to students working part-time.
Figure 1 Level of stress among categories of working students.
In figure 1, the median level of stress is the same across all categories of students, except students who work between 30-40 hours. So, this figure shows that when the students start working 30-40 hours then on average, they will have a higher level of stress. There are 20 students who does not work, but all of them are full-time students and 9 out of 20 students are taking more than 13 credits this semester. This explains the reason why the students who do not work to have a high level of stress, even when they do not have any stress from work. In addition, students who work 20-30 hours are all full-time students, which explains the reason they all have a high level of stress. This category of students has stress from school and work, which results in them expressing themselves as having a high level of stress.
Figure 2 Stress level among full-time and part-time student
In figure 2, we had only one part-time student and 48 full-time students. Surprisingly the median stress levels of full-time students was the same as that of the part-time student. The data above is not significant because we do not have enough part-time students to compare it with full-time student’s stress levels.
Figure 3 Stress level in male and female
In figure 3, although the number of females is twice of male, there is greater variation in stress levels among males compared to females. There are 35 females and 14 males, but according to the figure, 3 males have more variation in stress levels compared to females. They both have a median stress level of 4.
Figure 4 Level of stress in Religious and non-religious students
In figure 4, We observed that no participant declared themselves as having no stress, and most of the religious participants answered as having stress on the fourth scale, which is one step away from declaring oneself as extremely stressed. We had 26 participants and 23 participants declare themselves as religious, not religious respectively. Only one religious student answered as having a stress level of 2, which is one step away from being rarely stressed. We expected religious students to have low-stress level but none of the religious students told that they have low-level stress. All of them answered having a high level of stress, whereas we can see a huge variation on stress levels among the non-religious student.
Figure 5 Variation in age group found across different level of stress
In the above figure 5, we observe the presence of outlier for the stress level of 3. This happened because we only have one participant of age 22, and that participant happened to have a stress level of 3, which made it into an outlier. But in the case of stress level of 2, although we have a lot of students of age 18, only one of them reported having low stress.
Conclusion (Abdul Murshed)
In this study, we found that no participants reported feeling rarely stressed, but most of them reported being the fourth scale on the level of stress, and some of them answered as being extremely stressed. Although we found some observations, from the figures but the data we obtained from them are not significant. All our analyses showed that stress level is the same across all categories of age, gender, and full-time vs part-time students. Besides, when comparing all categories of working students, the level of stress they feel is similar across all working students, except for those working 30-40 hours, who had higher stress levels on average.
From, the non-parametric test we found that none of the data supports the hypothesis. This is due to lack of enough sample size. We can get better data if we have more questionnaire to measure the level of stress, instead of letting the participants decide for themselves. The data we obtained could be influenced due to the time when we were surveying because we had just finished mid-term and some classes had a third exam coming before the thanksgiving week. This may have caused the students to perceive having a higher level of stress.
Reference
Dawes, J. (2008). Do data characteristics change according to the number of scale points used? An experiment using 5-point, 7-point and 10-point scales. International journal of market research, 50(1), 61-104.
Beninato, J. P. (2012). Beliefs and Coping with Life Stress among UConn Students.
College students cite stress as key factor in academic performance. (2009, December). Clinical Psychiatry News, 37(12), 30. Retrieved from https://link-gale-com.ccny-proxy1.libr.ccny.cuny.edu/apps/doc/A215844769/HWRC?u=cuny_ccny&sid=HWRC&xid=df1191b7
Stress and your health: MedlinePlus Medical Encyclopedia. (2019, December 2). Retrieved from https://medlineplus.gov/ency/article/003211.htm.
Brougham, R. R., Zail, C. M., Mendoza, C. M., & Miller, J.R. (2009, February 11). Stress, Sex Differences, and Coping Strategies Among College Students. Retrieved from https://link.springer.com/article/10.1007/s12144-009-9047-0.
Appendix A
Stress level of CCNY students survey
- Age ________
- Gender ________
- Religious? Yes No
- International student? Yes No
- How many hours do you work? _________
- How many credits are you taking this semester? __________
On a scale of 1 to 5 how stressed do you feel? Stress refers to physical or emotional tension.
1 2 3 4 5
Rarely stressed O O O O O Extremely stressed
Tenjing Sherpa 