Socioeconomic Status & The Need to Belong:

An ANOVA Data Analysis

Elizabeth F. Logan

Department of Psychological Sciences, Texas Tech University

Research Methods 3401

Dr. Lindsay Greenlee

April 14, 2022

 ABSTRACT
The research study would like to examine the socioeconomic status of Texas Tech

students and how this may contribute to the feeling of needing to belong. The objective of this

research is to understand the various levels of socioeconomic status and how this can affect the

mental health of students at Texas Tech University. Using previous peer-reviewed scientific

works, I produced the hypotheses that support the meaningful differences between

socioeconomic status and need to belong using an analysis of variance (ANOVA) to measure the

question ratings based on a Likert scale measurement. The purpose of this study is to see if

there are any discrepancies that often hinder students due to socioeconomic status. To better

understand available treatments if there are major discrepancies within our findings.

Understanding how the data correlates with the research question and the limitations that may

have determined the distribution of the procedure.

INTRODUCTION
In the United States, seeking a higher-level education is a rite of passage for many high

school seniors. Over the last 40 years, we can visualize how the social part of college, the cost

of college, and the availability of college have changed drastically. Due to the higher cost of

college, more scholarships are available to students who have high academic success or who

need financial assistance. Many college students are enrolled at an institution with distinct

levels of socioeconomic status, because of this, there may be a cultural, social, or psychological
differences between individuals Universities have made practices to better suit the needs of all

students from a diversity of backgrounds. Socioeconomic status is an economic class standing

for individuals/groups that define their access to certain experiences or resources.

There are various levels of socioeconomic status in students that attend Texas Tech

University who have come from several types of backgrounds and resources. Although, there

can often be disparities on a college campus due to the lack of resources that students are able

to uphold due to a lower income background of that student. This often hinders the college

experience of that individual and can affect the students academically. Most of the time,

students who are struggling financially often find themselves working three part-time jobs, just

to be able to pay for their tuition and stay in school. This can also negatively affect their social

life with individuals because they do not have the same privilege as other individuals who have

more resources to balance their job, school, and social life.

The need to belong on college campuses is what many students want to achieve during

their time in university other than earning academic achievement. Due to the rising levels of

mental health awareness on college campuses after the emergence of the pandemic, social

interaction between students is critical to maintain a balanced and healthy college education.

The need to belong to college is important because of its benefits of having human interaction

with other peers and to also create a bond with like-minded people with similar goals in life.

Not only is the need to belong important for your social and emotional health, but it is crucial

for your academic success as well. Studies have shown that you are more likely to succeed in

college if you obtain healthy relationships and social life with other students. The good

outweighs the bad when you feel like you belong in college.
 Marginalization on college campuses often hinders the academic success of students

who have a lower socioeconomic background, eventually leading to the negative health and

well-being of those groups of individuals. There may be a relationship between being successful

in lower socioeconomic status, although a societal narrative of inequality or discrimination may

affect their success at university (Destin, et al., 2021).

Due to a diverse group of socioeconomic status on college campuses, there may be

distinct levels of self-esteem due to status and identity. The correlation between socioeconomic

status and self-esteem may cause distinct levels of sociometric status as well. Manipulating the

SES of individuals to be of higher socioeconomic status, caused the treatment of these

individuals to be treated better and therefore had better self-esteem. (Nikhila, et al., 2021)

Being a part of groups that are similar in socioeconomic status may affect the certainty

of an event or opinion that happens within that group, or what is also known as groupthink

theory. It was conducted that individuals who were certain about their orientation were more

likely to be more affected by their environment, which causes them to be led by biased claims

within groupthink. To prevent the negative outcomes of groupthink, it is known to be beneficial

to be a part of an environment that accepts open-ended discussions that lessen the outcome of

biased opinions (Hodson, et al., 1997).

Due to the negative effect of the COVID-19 pandemic, college students' sense of

belonging to their college campus has impacted their mental health and well-being of students.

The practice of isolation led to decreased rates of essential social interactions, causing the rates

of anxiety and depression to increase among students of different demographics. Before and
after the effects of the pandemic, underrepresented groups of individuals were reported to be

less likely to belong in a group compared to other populations. Due to the relation between a

sense of belonging and mental health, the importance of being in inclusive and positive

environments that accept human interaction is essential (Gopalan, et al., 2022).

Lower socioeconomic status in students who go to university may experience limited

access to resources which may affect their sense of belonging during their time in university.

Due to students having higher rates or lower rates of prestige due to socioeconomic status,

research shows that individuals with lower SES experience low levels of prestige, whereas

individuals with higher SES experience prominent levels of prestige. The correlation shows how

the amount of prestige in SES groups may affect a sense of belonging in college. (Jury, et al.,

2019)

The research studies represent various connections and the relevance of both the need

to belong and socioeconomic status, and that these variables would be something to research

on in this study. Therefore, the goal of this study is to be able to identify any concerning

differences between each socioeconomic status with the need to belong. Our independent

variable is the various levels of socioeconomic status in each participant and the dependent

variable is the need to belong.

I hypothesize that there will be many individuals who will be affected by need to belong

due to their level of socioeconomic status, because of resources and the balance between

getting an education that will benefit them financially or going out and meeting other

individuals who support you due to a certain level of socioeconomic status. I predict that there
will be outliers that create discrepancies within the dataset and that there will be many

differences between each socioeconomic status on their need to belong, which will then reflect

the nature and relevance between socioeconomic status and the need to belong. Therefore, is

there a correlation between college students of socioeconomic status and needing to belong?

METHODOLOGY

Participants

In this study there were a total of 775 participants: race identity = 772 participants,

gender = 768 participants, and then sexual orientation = 767 participants that were valid in the

data set. For gender: Males = 30.5%, Females = 66.1%, Nonbinaries = 1.7%, Preferred not to say

= 0.8%.

For ethnicity/race identity: Black/African American = 5.8%, Hispanic/Latino American =

24.3%, East Asian = 0.9%, South Asian = 1.4%, Middle Eastern = 0.9%, Native American = 0.1%,

Caucasian/White = 63.9%, Multiracial = 1.5%, and another race not identified = 0.3%.

For sexual orientation: Bisexual = 11.5%, Gay/Lesbian = 3.6%, Heterosexual = 81.9%,

Pansexual = 0.8%, Questioning/Unsure = 0.3%, Asexual = 0.4%, and Something not listed = 0.5%

For age, there was a valid listing of 697, where the minimum age was 18 and the

maximum age was 68. The age of the participants ended up being young (M = 22.73, SD = 7.68).

Therefore, the average age in the dataset was 22.73 years old.
Materials

In this study, we used the Likert scale to measure the need to belong, where the

participants answered questions on the survey based on a scale from 0 to 5, from strongly not

feeling belonged to strongly feeling belonged. A Likert-type scale involves a series of statements

that participants will choose to distinctively rate their responses on survey questions. We then

use a program called SPSS, where the data is then organized to identify the ANOVA, effect size

of the data, and examine the post-Hoc analysis so that the data will be examined properly.

Procedure

The participants were measured accordingly in the experiment using the Likert scale of

need to belong. After receiving informed consent from participants, Texas Tech students were

asked to answer a survey that included a link to the website where participants were then able

to answer the questions to the best of their ability.

RESULTS

ANOVA Column1 Column2 Column3 Column4 Column5

NTBtotal
Sum of Mean
Squares df Square F Sig.
Between
Groups 863.316 6 143.886 3.259 0.004
 Within Groups 33688.997 763 44.153
Total 34552.313 769

The ANOVA of the Between Groups in the Need to Belong total output has a significant

level of 0.004. Since the significant value of 0.004 is lesser than the alpha value of 0.05, the data

set is statistically significant. There is a difference between the socioeconomic status of college

students and need to belong. Assuming that the between groups had no effect, we would then

obtain the observed difference or more in the 4% of ANOVA tests are due to random sampling

error. Therefore, since p-value=0.004, the data is unlikely with a true null hypothesis

ANOVA Effect Colum Colu Colu Colu Colu

Sizesa,b Column1 n2 Column3 mn4 mn5 mn6 mn7
Point 95%
Estimat Confidence
e Interval
Uppe
Lower r
NTBtotal Eta-squared 0.025 0.003 0.043
Epsilon-squared 0.017 -0.005 0.036
Omega-squared
Fixed-effect 0.017 -0.005 0.036
Omega-squared
Random-effect 0.003 -0.001 0.006
a Eta-squared and
Epsilon-squared are
estimated based on
the fixed-effect
model.
b Negative but less
biased estimates
are retained, not
rounded to zero.
 The effect size analysis, there are three categories: small= 0.01, medium= 0.06, and

large= 0.14. In the ANOVA effect size data set, the Eta-squared analysis of n^2= 0.025, indicates

a small effect size.

Multiple Colu Colu Colum Colu

Comparisons Column1 Column2 mn3 mn4 Column5 n6 mn7
Dependent Variable:
NTBtotal
Bonferroni
Mean 95%
(J) Difference Std. Confidence
(I) classback classback (I-J) Error Sig. Interval
Upper
Lower Bound Bound
Middle 0.800
Lower-middle class class -1.55734 54 1 -3.9976 0.8829
Upper 1.354
class -0.07865 28 1 -4.2069 4.0496
Upper-
middle 0.818
class -1.21983 08 1 -3.7136 1.2739
Working 1.013
class 1.71653 94 1 -1.3743 4.8073
3.900 12.811
None 0.92135 5 1 -10.9685 2
Lower 4.751
class/poor -5.07865 08 1 -19.5613 9.404
Lower-
middle 0.800
Middle class class 1.55734 54 1 -0.8829 3.9976
Upper 1.217
class 1.47869 68 1 -2.2332 5.1905
Upper-
middle 0.563
class 0.33751 84 1 -1.3812 2.0563
 Working 0.822
class 3.27387* 64 0.002 0.7662 5.7815
3.855 14.230
None 2.47869 2 1 -9.2731 4
Lower 4.713 10.848
class/poor -3.52131 96 1 -17.8908 2
Lower-
middle 1.354
Upper class class 0.07865 28 1 -4.0496 4.2069
Middle 1.217
class -1.47869 68 1 -5.1905 2.2332
Upper-
middle 1.229
class -1.14118 28 1 -4.8884 2.606
Working 1.367
class 1.79518 46 1 -2.3732 5.9636
4.006 13.214
None 1 96 1 -11.2144 4
Lower 4.838
class/poor -5 87 1 -19.7503 9.7503
Lower-
middle 0.818
Upper-middle class class 1.21983 08 1 -1.2739 3.7136
Middle 0.563
class -0.33751 84 1 -2.0563 1.3812
Upper 1.229
class 1.14118 28 1 -2.606 4.8884
Working 0.839
class 2.93636* 71 0.01 0.3767 5.496
3.858 13.904
None 2.14118 88 1 -9.6218 1
Lower 4.716 10.519
class/poor -3.85882 97 1 -18.2375 9
Lower-
middle 1.013
Working class class -1.71653 94 1 -4.8073 1.3743
Middle 0.822 -
class -3.27387* 64 0.002 -5.7815 0.7662
Upper 1.367
class -1.79518 46 1 -5.9636 2.3732
Upper-
middle 0.839 -
class -2.93636* 71 0.01 -5.496 0.3767
 3.905 11.108
None -0.79518 09 1 -12.699 7
Lower 4.754
class/poor -6.79518 85 1 -21.2893 7.699
Lower-
middle 3.900 10.968
None class -0.92135 5 1 -12.8112 5
Middle 3.855
class -2.47869 2 1 -14.2304 9.2731
Upper 4.006 11.214
class -1 96 1 -13.2144 4
Upper-
middle 3.858
class -2.14118 88 1 -13.9041 9.6218
Working 3.905
class 0.79518 09 1 -11.1087 12.699
Lower 6.065 12.490
class/poor -6 84 1 -24.4904 4
Lower-
middle 4.751 19.561
Lower class/poor class 5.07865 08 1 -9.404 3
Middle 4.713 17.890
class 3.52131 96 1 -10.8482 8
Upper 4.838 19.750
class 5 87 1 -9.7503 3
Upper-
middle 4.716 18.237
class 3.85882 97 1 -10.5199 5
Working 4.754 21.289
class 6.79518 85 1 -7.699 3
6.065 24.490
None 6 84 1 -12.4904 4
* The mean
difference is
significant at the
0.05 level.

The mean difference in the above data set represents a statistically significant outcome

within the socioeconomic status and need to belong. The working class was statistically

significant among the upper-middle class and middle class. The upper-middle and middle
classes are both the same regarding the dataset due to the p-value being greater than 0.05

alpha. Since the other mean differences of the post-hoc analysis displays a significance of 1,

which is well above the alpha of 0.05, means that the data is not statistically significant and

should not matter when measuring differences in ANOVA. The mean differences between each

variable compared indicate that there is a correlation between socioeconomic status and the

need to belong. Because there are only a few meaningful differences between the variables

listed above and the effect size of the mean differences is small, the correlation between these

variables is not large.

DISCUSSION

The results of the data presented in each table indicate a small correlation between the

socioeconomic statuses of college students and the need to belong. Because the effect sizes are

small and only a few mean differences between each socioeconomic class are statistically

significant, tells us that there is not a strong correlation or a large correlation between the

ANOVA meta-analysis of socioeconomic status and need to belong. Comparing need to belong

with an ANOVA analysis of the working, middle, and upper-middle classes, it is observed that

due to small effect size and the selection size of the population does not have enough

variability, there is only a small difference. Because the difference is not large, the results do

not need immediate treatment.

 Again, is there a difference between college students of socioeconomic status and

needing to belong? I have found meaningful differences between three groups in my ANOVA

dataset that contributes to my original hypothesis.

The results that were collected within the dataset will be compared to other previous

scientific works and discussions that were previously mentioned. Many college students are

enrolled at an institution with distinct levels of socioeconomic status, because of this, there

may be a cultural, social, or psychological difference between individuals. Universities have

conducted treatments to better suit the needs of all students from diverse backgrounds.

According to Destin’s hypothesis and scientific assumption: marginalization on college

campuses often hinders the academic success of students who have a lower socioeconomic

background, eventually leading to the negative health and well-being of those groups of

individuals. There may be a relationship between being successful in lower socioeconomic

status, although a societal narrative of inequality or discrimination may affect their success at

university (Destin, et al., 2021). The difference between the study conducted by Destin and this

study, is that in the data set, there were no major differences to determine lower

socioeconomic students are hindered from academic success. It can be analyzed that there is a

surprising outcome within the dataset that does not reflect the outcome of Destin’s hypothesis.

According to Nikhila’s on self-esteem in college students with a lower socioeconomic

status: due to a diverse group of socioeconomic status, there may be a diverse group of self-

esteem due to status and identity. The correlation between socioeconomic status and self-

esteem may cause distinct levels of sociometric status as well. Manipulating the SES of
individuals to be higher caused treatment of these individuals to be treated better and have

better self-esteem. (Nikhila, et al., 2021) The difference between Nikhila’s correlation and this

current ANOVA difference of variables, is that we are using an ANOVA analysis that measures

only differences between groups, not using a correlation to measure these outcomes. Nikhila’s

study compares two means to find a direct outcome whereas our study measures the

differences that may determine a higher or lower outcome of need to belong.

As we compare the hypotheses from each source to our dataset, the interpretations

that experimenters used to hypothesize their correlations may have carried over into the three

independent variables that were identified in the dataset, but not in a profound sense of

urgency to provide immediate treatment. Because of the outcome of the effect size and meta-

analysis, there was not a significant difference between the three independent variables and

the dependent variable to confirm strong internal validity.

To better understand the comparisons of literary works of other studies to the study

that was conducted for my research question, we must analyze the internal validity of the

results, understanding that the following results may not have a strong internal validity. This

can be due to longer than usual questions on the survey or the relevance of categorical or

continuous measurements of participants in a randomized population. The bigger the

population that participated in the study, the stronger the internal validity is.

Limitations of this study can be the validity and reliability of the results that were

collected. If the results were to consist of answers that were completely inaccurate to the effect

size and the meta-analysis, internal validity will be at risk. Since there were 700 students who
participated in the survey, there is a higher chance for internal validity and for external validity.

Internal reliability without internal validity will be an extremely risky and poorly designed study.

Without the right methodology for internal validity, there is no reliability in the study at all.

Another limitation in the study can also consist of errors in the data set. This is because of the

environment in which the student may have answered the survey questions. Due to the

expensive cost of college, there may have been limitations on lower-class students who may

not have access to the same resources to answer survey questions. The data may have been

skewed from an environment that was distracting to the participant. Therefore, the student

may not have answered all the questions in a valid manner. A sampling error of the population

may affect the data, due to the lack of experimental design or from rushed experimentation

where the data was not measured at its full scale. Other limitations can include influences of

other categories such as discrimination in the dataset that excludes people who identify as a

specific gender, sexual orientation, or race and ethnicity.

In future experimentation, there can be a better knowledge of certain research methods

that will be most appropriate for participants to answer on surveys with a type of hypothesis

testing you want to accomplish. This can be an ANOVA test, t-test, or one-sided hypothesis test.

Other ways of conducting future experiments are organizing the procedures and materials that

you will be using, specifying an alternative hypothesis other than the null hypothesis, a broad

understanding of other scientific studies to expand your knowledge about APA research topics,

and having a better understanding of the environment that participants are brought into when

conducting your study. What researchers should consider for future experiments is being able

to conduct a survey or experiment with a large population sample to increase sample validity.
These changes in future experimentation can better accommodate the survey results that are

unfiltered and true to its nature.

Overall, it is important to note that experimentation in psychology often takes years of

practice within the field to gain an understanding of research methods. Therefore, future

practices in research can affect society as a whole because of new discoveries that are being

made by every research question that you can produce. Asking questions and using the

scientific method makes a difference in the validity of the study and the overall quality of a

study. There will always be statistical errors and unexpected outcomes that are a result from

conducting the study, that is why experimentation must be done to answer the unknowns

about science.

REFERENCES

1. Destin, M., Rosario, R. J., & Vossoughi, S. (2021). Elevating the objectives of higher education

to effectively serve students from diverse socioeconomic backgrounds. Policy Insights from the

Behavioral and Brain Sciences, 8(1), 59–66.

2. Mahadevan, N., Gregg, A. P., & Sedikides, C. (2021). Self-esteem as a hierometer: Sociometric

status is a more potent and proximate predictor of self-esteem than socioeconomic status.

Journal of Experimental Psychology: General, 150(12), 2613–2635.

3. Hodson, G., & Sorrentino, R. M. (1997). Groupthink and uncertainty orientation: Personality

differences in reactivity to the group situation. Group Dynamics: Theory, Research, and

Practice, 1(2), 144–155.

4. Gopalan, M., Linden-Carmichael, A., & Lanza, S. (2022). College students’ sense of belonging

and mental health amidst the COVID-19 pandemic. Journal of Adolescent Health, 70(2), 228–

233.

5. Jury, M., Aelenei, C., Chen, C., Darnon, C., & Elliot, A. J. (2019). Examining the role of

perceived prestige in the link between students’ subjective socioeconomic status and sense of

belonging. Group Processes & Intergroup Relations, 22(3), 356–370.