The Emotional Resiliency Scale on Transition of Learning Modality

In Partial Fulfillment of the Requirements

in Psychological Test Development

December 15, 2022

 INTRODUCTION

I. Rationale

An individual's capacity to handle pressure and recover from challenging and

personal circumstances is primarily determined by emotional resilience, a cornerstone

of mental well-being and positive psychology. According to Hurley and Young (2022),

a person needs skills and experience to develop resilience. Furthermore, they may

encounter obstacles because it requires time, effort, and support from the people around

them. However, it relies on their actions and capabilities (such as self-confidence and

conversational skills) and outside factors (like social support and available resources).

In addition, numerous academic opportunities for students are wasted during the peak

of COVID - 19. Since there are health protocols that must be followed, the educational

system was forced to utilize the advancement of technology and slowly adapt to the

new normal changes for learning set-up. However, students are still adjusting to the

new learning environment and transitioning from virtual to face-to-face classes. It

appears that the epidemic has likely worsened inequality in students' accessibility to

educational materials, the development of social connections, and the development of

their sense of self. Additionally, these discrepancies will influence more than simply

academic results.

In relation to the current academic set-up, the proponents developed a reliable

and valid scale to measure students’ emotional resilience regarding the transition in

learning modality. This scale can be used as a screening tool for every school and

university to know if the student is undergoing emotional challenges and help the

student cope with their problems to equip them for their future professions.
 II. Definition of Terms

Emotional Resilience

It is described as the ability to adapt positively to environmental demands

following stressful situations and major crises (Gross, 2017). The ability of an

individual to manage stress and control their reaction and behavior under negative

experiences. It is the innate motivation, a power within ourselves, that enables us to

endure and overcome life's challenges.

Relationships

It is described as an interaction between close friends, peers, family, or adults

within the community. It provides support and influences how emotionally resilient a

person can be in the face of a negative experience. August & Rook (2013) also refer to

relationships as connections between people with recurring interactions that the

participants perceive to have personal meaning.

Identity

It is described as a personal and a collective sense of purpose, self-appraisal of

strengths and weaknesses, aspirations, beliefs, and values, including spiritual and

religious identification. It is the way we position ourselves within a society as well as

in the world (Woodward, 1997)

III. Theoretical Framework

Michael Ungar's resilience theory is the most suitable theory for the proponent's

developed scale. The subconstructs of the theory will be a basis to create numerous

items that would evaluate students' emotional resilience in terms of the educational

system's shift from digital learning to face-to-face classes. According to Ungar's theory,

resilience may increase organizational and individual adaptation in times of crisis.

Many organizations have integrated this idea of resilience as a functional procedure that

enhances safety and social responsibilities. Furthermore, according to his hypothesis,

when resources are abundant, individuals are more inclined to practice and exhibit

resilience. The environment may either stimulate or repress a person's motivation, sense

of power, attitudes, personality characteristics, and biological predispositions for

specific behaviors, such as recklessness and anxiousness. In general, Ungar's theory of

resilience highlights the features of people and their environments that promote

resilience.

Consequently, Michael Ungar's resilience theory has seven (7) tensions: access

to material resources, relationships, identity, power and control, social justice, cultural

adherence, and cohesion. However, with the proponent's developed scale, two (2) sub-

constructs are utilized: relationship and identity. Moreover, relationships strongly

emphasize ties to relatives, close friends, and colleagues, while identity emphasizes a

person's skills, shortcomings, and a feeling of collective purpose. Furthermore, the

measurement of academic pressure and stressful events will be based on these two (2)

tensions.

METHODS AND STATISTICAL RESULT

I. Test Development Process

1. Test Conceptualization

In psychology, emotional resilience is the capacity to generate positive emotions

and bounce back from difficult emotional situations. It has two fundamental qualities:

the ability to evoke positive emotions in the face of unpleasant emotional stimuli and

the capacity to bounce back from unpleasant emotional experiences (Davidson, 2001;

Conway & McDonough, 2006). Resilience has been chiefly researched as a protective
factor in children undergoing substantial life transitions, adversity, and stress (Haggerty

et al. 1996). Resilience can support individuals in handling transition-related

challenges, adjusting to change, and coping with uncertainty in the context of education.

For students to successfully handle emotionally taxing situations, develop proper

coping mechanisms, improve well-being, and progress professionally, emotional

resilience may be essential.

In relation to the person's emotional efforts in stressful situations, emotional

resilience is more likely to result in emotionally directed coping. This adaptive process

recovers and sustains normal emotional levels in the face of catastrophic experiences.

Adaptation can result from interactions between an individual and their environment.

The individual's mental health and capacity for social adjustment are significantly

influenced by emotional resilience. Differences in emotional resilience may cause

individual differences in emotional responses, and these differences may then affect

how cognitive processes deal with dynamic input (Zhang & Lu, 2010).

The coronavirus (COVID-19) has spread worldwide and impacted many facets

of society, including higher education, which has not been spared. Due to the pandemic,

education has recently changed all across the world. Schools were forced to conduct

online classes as a precaution against the increasing cases of people getting infected by

the virus. In a systematic analysis conducted by Talib et al. (2021), the shift from in-

person instruction to online distance learning, the impact of lockdowns and other

pandemic-related actions on students' personal and academic lives, and their

experiences with online learning. The researchers also noted that the pandemic itself

might catalyze change, but they cautioned that there were potentially serious issues that

needed to be addressed, such as disparities and problems with accessibility,

insufficiency, poor communication, technological difficulties, the requirement for

digital literacy, school and personal life balance, worries about privacy in online

settings, challenges with evaluating student achievement, and a lack of resources.

In another study, Aristovnik et al. (2020) examined the COVID-19 pandemic's

global effects on the lives of higher education students. They claimed that the dramatic

adjustments brought about by the measures put in place to solve COVID-19

significantly influenced students' emotional well-being in higher education, adversely

affecting their academic performance. They said that the fact that the epidemic began

in the southern hemisphere at the start of the academic year while it started in the

northern hemisphere in the middle of the academic year caused the two hemispheres to

react emotionally in different ways.

At present, life gradually returned to normality after the pandemic, which

appears to have stopped; for the time being, consideration must be given to the return

to in-person instruction and its ramifications. Therefore, the proponents aim to develop

a psychometrically valid and reliable test to determine students’ emotional resilience in

their transition from online learning to face-to-face education.

The primary target of the scale will be the pre-med students under the Institute

of Arts and Sciences (BS Psychology & BS Biology) and Institute of Health Sciences

and Nursing (BS Medical Technology & BS Nursing). The proponents developed an

emotional resilience scale with three sub-constructs: relationships, identity, and

cohesion. Further, the test developers designed this test to determine the emotional

resilience of the target individual based on how high or low his final score will be. In

that regard, the current literature states that the return to face-to-face instruction needs

attention. The test being developed can be utilized to measure if emotional resiliency

and environmental factors influence how they can react and cope with the transition
back to school classes in this pandemic. Additionally, it will raise awareness of the

effects of COVID-19 on the student’s academic performance and coping mechanisms

during this transition to their college.

2. Test Construction

a. Scaling

To measure the students’ emotional resilience, the proponents have chosen a

dichotomous scale. According to Stöber, Dette, & Musch (2002), dichotomous scoring

confounds the tendency of participants to provide extreme and socially desirable

responses. Since the dichotomous scale provides two answer options, the proponents

would get a direct and precise answer and avoid vague and ambiguous responses.

However, scales with two points often perform poorly than scales with more categories

(Lundmark et al., 2016 On the contrary, Krosnick (2005) stated that rating scales need

more interpretive work than dichotomous scales, which can compromise the

consistency of the responses. Additionally, Alwin (2007) also claims that absolute

metric scales and rating scales both have lower reliability than dichotomous scales.

Hence, a dichotomous scale was utilized to measure participants’ agreement for each

statement. The participants can choose from “Yes” or “No” to answer each item.

b. Item Writing

The total item created for the item pool is 55, with 20 items for the relationship

subconstruct, 20 for the identity subconstruct, and 15 for the cohesion subconstruct.

One item is subjected to reverse scoring. The item poll was submitted to three expert

validators for evaluation. After the panel validation, where items have been examined

for their content, the total number of items included in the updated version of the item

pool consists of 40 items. Subsequently, the proponents have chosen to remove the
cohesion subconstruct in line with the suggestion of the expert validators. Since a

dichotomous scale was utilized where participants could choose their answers from a

given set of responses, the items were categorized in a selected-response format. After

the evaluation process, the final version of the test consists of 29 normal-scored items

and 1 reverse-scored item, with a total of 30 items: 15 items for the relationship

subconstruct and 15 items for the identity subconstruct.

c. Scoring

Since a dichotomous scale was employed to statistically measure the

participants’ responses, the score ranged from 1 (lowest) to 2 (highest), which are the

following: 1 = No, 2 = Yes. The participant may select an answer to indicate if the

statement is applied to them. Subsequently, the score interpretation for the scale will be

based on the participant’s average cumulative total scores for each item.

3. Test Tryout

To administer the scale to the desired participants, the proponents used both

Google forms and the pen-and-paper method to gather the participants efficiently. The

data was collected from 150 participants at Far Eastern University Manila. The scale

developers posted publication materials to different social media accounts introducing

the scale and the qualifications of the participants. Before answering the scale, the

participants will be asked for their personal information and informed consent

indicating what the scale is all about, who the researchers are, how their information

will be stored, their confidentiality and anonymity, and their right to withdraw from the

scale. The volunteered participants will first answer the Emotional Resilience Scale,

which has 30 items made by the proponents. After that, in the same form, they will also

answer the concurrent scale of the Resilience Scale for Adolescents of Hjemdal et al.

(2006). Furthermore, the personal information and data gathered from the participants
 will be stored in a secured Google Sheet that will only be made available to all the

proponents.

4. Item Analysis

In analyzing the dichotomous scale items, emotional resilience will be

determined in two sub-constructs, namely, “Relationship” and “Identity”. The score 2,

equivalent to “Yes” will be interpreted as high, while 1, equivalent to “No” will be

interpreted as low. Participants who responded favorably on the items with normal

scoring will be interpreted as having a high level of emotional resilience during the

learning modality transition. In contrast, a low score would indicate that the participant

has low emotional resilience. High scores on the items designed for reverse scoring

would suggest that the participants have low emotional resilience. On the contrary,

scoring low will be interpreted as having a high emotional resilience during the learning

modality transition among the participants.

5. Revisions and Recommendations

Figure 1 shows the Emotional Resiliency Scale on the Transition of Learning

Modality used for data gathering

SUBCOMPONENT # 1: RELATIONSHIP YES NO

1. I can get support from my family when I'm under too much
pressure because of the transition in the learning environment

2. I receive support from my friends when there is too much

pressure from my studies.

3. I can discuss my academic challenges with my friends.

4. I have my family who would listen to my complaints about the

school.

5. I receive encouragement from my family and friends.

6. I can seek help from my professor when I’m having difficulty
with my studies because we can meet during face-to-face
classes.

7. I do not have anyone to rely on during emergencies.

8. I can refocus and improve my academic performance with the

help of my family and friends.

9. I can overcome academic challenges with the help of my

family.

10. I can finish school activities efficiently with the guidance of

my friends and classmates.

11. I find it hard to make friends in face-to-face classes due to

health and safety protocols.

12. I feel nervous and worried whenever my friends, classmates,

and professors get the Covid-19 virus while face-to-face
classes are ongoing.

13. I have classmates whom I can rely on when I need help with
my studies.

14. I can easily adapt to the new learning environment because I

have my friends with me.

15. When in doubt, my friends gave me advice to remain calm and

believe in my capabilities.

16. I can get assistance from our guidance counselor to help me

with the struggles of adjusting to the new learning
environment.

17. I feel more enthusiastic about joining group work since it is

more accessible to communicate with them.
18. When I feel exhausted from schoolwork, I try meditation and
join physical activities with my friends.

19. My family keeps me motivated despite my difficulties with the

transition of learning modality.

20. I feel happy when professors give feedback on my efforts in

understanding a lesson despite the transition in learning
modality.

SUBCOMPONENT # 2: IDENTITY YES NO

1. I am aware of the strengths and weaknesses that I have during

the transition of modality.

2. I have strong beliefs that I can overcome all the struggles

during the transition to face-to-face classes.

3. I am able to adapt easily to changes in learning modalities.

4. I remain hopeful that I can achieve my goals even after the

transition of learning modalities.

5. I have confidence in my ability to complete tasks, even with

the adjustments of modalities.

6. I have the capacity to cope with negative or painful emotions

like guilt, fear, and sadness during the transition of modality.

7. My self-belief is affected when a task doesn't go as planned

during the change of modality.

8. The shift in learning modes makes me doubt my commitment

to my studies.

9. I can recover quickly from illnesses, injuries, and other

setbacks during the transition of learning modalities.

10. During the shifting of modalities, I feel like asking for help
implies weakness.
11. Even under the pressure of studying and the transition of
modalities, I remain focused and think clearly.

12. I firmly think that God has a purpose for everything that is
happening right now and during the shifting of modality.

13. Even when hopeless due to academic pressure and the

transition of modes of learning, I do not give up easily.

14. I have strategies in place to manage my stress during the

transition of modes of learning.

15. I'm able to overcome unpleasant memories and lessen their

influence on my emotions and actions during the transition of
modality.

16. I believe that God helps me to cope with all the problems that I
encounter during the shifting of modality.

17. I find it difficult to recall my accomplishments, and failures are

difficult for me to forget during the shifting of modes of
learning.

18. During the transition of modality, I believe that having to cope

with academic pressure and stress can make me stronger.

19. I strive to improve myself academically so that I can adapt to

the change in learning modes swiftly.

20. I use feedback and past academic achievements to motivate

myself in studying during the transition of modes of learning.

SUBCOMPONENT # 3: COHESION YES NO

1. I have time to do my hobbies during the pandemic.

2. I started spending less time bonding with my family because of

the transition to face-to-face classes.
3. I can balance my academic responsibilities and household
responsibilities during online classes.

4. I feel safe knowing that my friends understand my struggle in

transitioning from online set-up to face-to-face classes.

5. I have a strong support system that allows me to show my

frustrations about learning modality transitions.

6. I am glad that my family allows me to have personal space

during online classes.

7. I have an amazing social environment that allows me to utilize

my time as a student during online classes.

8. I have a strong support system that allows me to strive in my

own phase, both during online classes and my transition to a
face-to-face class.

9. I am having a hard time establishing boundaries when I am

around my family during online classes.

10. The people around me invalidate my struggles during face-to-

face class.

11. My parents provided my basic needs during online classes

through the transition to face-to-face classes, which helped me
accomplish my academic requirements.

12. I still have time to pursue my personal interest while striving

for academic success.

13. The people around me ignore my academic struggle during my

learning modality transition, making me feel worthless.

14. I have a productive coping mechanism that allows me to

bounce back from feelings of defeat during face-to-face
classes.
15. The people around me make me feel valued even when I
struggle academically, especially during the learning modality
transition.

After consulting with three validators from the faculty of Psychology of FEU,

the proponents revised the test items. The original emotional resiliency scale included

twenty questions for sub-construct one and two (relationship and identity) and fifteen

questions for sub-construct three (cohesion); there were fifty-five items. However,

based on the validators' accepted, rejected, and suggestions for revision of items, it was

condensed into thirty items in total, with fifteen questions for both sub-constructs one

and two (relationship and identity). Therefore, with the suggestions of the three

validators,' the proponents decided not to include the third subcontract, which is

cohesion, since it overlaps with the two sub-constructs.

Recommendations:

As for the recommendations, the proponents recommend to future test

developers who would explore the same construct as emotional resiliency to use the

Likert scale instead of the dichotomous scale. The Likert scale would give the

participants a more comprehensive range of options that best correspond with their

feelings about the statement or question.

Another recommendation is to include more sub-constructs under the seven

(7) tensions of Ungar's theory of resilience. The additional sub-constructs would give

more test items to measure the participants' emotional resiliency, which can provide

better results.

The proponents also recommend using other theories that would allow them

to explore other facets of emotional resilience in terms of transition, which would allow

for a more comprehensive analysis of the subject.

 Regarding the participants, the proponents recommend reconsidering the

target participants' qualifications. Since target participants of the developed scale

were limited to students taking pre-med courses, including other courses for a wide

range population would give the study a more generalized sample.

In line with the suggested result of the reliability, item number twenty-two (22)

under identity sub-construct must be reverse scored to make it positively correlated to

other items.

Lastly, the proponents recommend finding another test for concurrent validity

since the chosen standardized test is intended for adolescents, which is not appropriate

for the age group of the target participants.

II. Standardization Sample

Figure 2 shows the Mean and SD calculated using Jamovi from data

acquired during the test administration.

Figure 3 shows the descriptive statistics calculated using JAMOVI.

 The gathered data of the proponents from test administration indicates that

overall, the mean is 1.23 and the standard deviation is 0.166 (see figure 2). The standard

deviation represents how dispersed the values are around the mean in a data set. The

low SD proves that the answers gathered from the participants are close to the mean.

In the gathered data of the proponents, the obtained value of the standard deviation is

close to zero. It does not exceed a ±2SD, representing data gathered is closer to the true

value than those who fall outside the ±2SD.

Figure 3 consists of the items showing an almost uniform mean point of 1, not

exceedingly more than 1.57 (see figure 3). The proponents' dichotomous choices “YES

or NO” lead to the mean result that is not deviating from each other. The proponents

correspond “YES” as number 2 and “NO” as number 1. The proponents' RQ and IQ

dimensions represent the Relationship Questions and Identity Questions domains. From

the original test concept, the domain of Cohesion was included; however, upon

revisions, the proponents found it best to divide the approved components from
Cohesion to RQ and IQ to avoid redundancy of components per question (see figure 1).

The corresponding number of the dichotomous items shows that most participants

answered “NO” to both RQ and IQ.

The skewness of the results lies in the middle part of the distribution, with a

mean of 1.23, thus making it positively skewed. The Standard Error, which is the

amount of discrepancy that can be expected in a sample estimate compared to the value

of that population, is 0.198, which is a relatively acceptable SE. The kurtosis (see figure

3) are scattered, and there is no definite part where the answer of the respondents match

the same peak. However, the kurtosis shows most positive values and can be considered

under leptokurtic distribution with a SE of 0.394.

III. Reliability

After using Jamovi for data analysis, the Cronbach alpha of the gathered data is

0.836, which is a relatively good reliability value. Thus the proponents can conclude

that the test is reliable and consistent. The reliability of the test is also supported by the

results showing that almost all 30 items show a positive relationship and are

significantly correlated except for item IQ22. The latter shows that it correlates

negatively with the total scale and probably should be reversed.

IV. Validity

● Content Validity

To guarantee that content validity could be clearly established, two methods

were implemented. The test developers initially conceived the content domain of the

construct by outlining its components and dimensions before creating test items. In

order to determine the specific area that the items will rely on, the precise definition

and properties of the subcomponents were further reviewed. In accordance with

Ungar's Resilience Theory, the test items were subsequently constructed based on the

two references of the subcomponents following.

Relationship. Relationships with close friends, peers, family, or adults within

the community provide support and influence how emotionally resilient a person can

be in the face of negative experiences.

Identity. Personal and collective senses of purpose, self-appraisal of strengths

and weaknesses, aspirations, beliefs, and values, including spiritual and religious

identification.

Initially, the test developers had intended to develop a total of 55 items for the

scale, of which 20 and 35 were created for the relationship and identity subcomponents.

However, they eventually decided to employ only 30, based on the items that were

subjected to be administered based on the psychometricians. On the 55-item scale, 33

were approved, 15 were rejected, and 7 were given revision recommendations.

● Concurrent Validity

The degree of synchrony between 2 distinct assessments is called concurrent

validity. In general, one assessment is relatively new, while the other is well-established

and has already been shown to be reliable (Adams et al., 2014). Consequently, the

Resilience Scale for Adolescents was a guide for the test developers. The general public

is the intended audience for this resilience scale that is continuously in use.

Furthermore, the already established resilience scale for adolescents utilized a

5-point Likert Scale with a rating scale ranging from 1 (strongly disagree) to 5 (strongly

agree). It consists of 28 items, and the original scale is composed of five factors: (1)

Structured Style, (2) Social Competence, (3) Personal Competence, (4) Social

Resources, and (5) Family Cohesion. It retained the conceptual content from its
predecessor to assess adult resilience. Additionally, Cronbach's alpha was used to

calculate the subconstructs' reliability indices, which are as follows: Personal

Competence (=.85), Social Competence (=.82), Structured Style (=.69), Family

Cohesion (=.85), and Social Resources (=.78). Confirmatory factor analysis was used

to evaluate the construct validity of the original measure. Using the statistical package

AMOS, the confirmatory factor analysis was carried out using a composite model that

accounted for the relationships between the five factors identified in the exploratory

factor analysis. Items with the lowest factorial standardized saturation were excluded

after the analysis. The reliability indices in both measures were enhanced by removing

those two items, which led to Cronbach's Alpha values of.842 for family cohesion and

.710 for social resources.

The result suggests that it has a value of -0.595 which means a negative

correlation between the developed scale and concurrent scale because they are not

similar in terms of the total number of items and scale format.

● Construct Validity

Proving a method's overall validity is important. The way a test evaluates the

concept it was intended to evaluate is known as its construct validity (Bhandari, 2022).

One of the most effective methods, factor analysis, will be used to prove this validity.

Utilizing statistical techniques, factor analysis enables us to investigate the underlying

dimensions that account for the interactions between a number of complex items

(Tavakol & Wetzel, 2020). It is now simple to discover and comprehend the

relationships between the numerous items of the constructed test and its underlying

components that may share similarities because the table that displays the correlation

coefficients for various variables has been simplified. In order to determine the

variables that load concurrently and measure the same component, it is possible to

evaluate the evidence based on the developed test's internal structure as well as

evidence-based on its content. This allows for the analysis of both evidence based on

the constructed test's content and evidence based on its internal structure to identify

factors that load concurrently and assess the same component.

 V. Factor Analysis

Table 1

Correlation Matrix
 Test developers used the correlation matrix, which provides a uniform

perspective of all correlations, to comprehend how connected all the variables are.

Briefly explained, a correlation matrix is a table with correlation coefficients for your

variables (Krishnan, 2021). The correlation coefficient for various variables is shown

in this table. It reveals that 18 items across the scale's two subcomponents are more than

0.80 (>0.80), which indicates that those test items should be excluded from the scale

test. As a result, only 12 test items from the constructed test are maintained.

Table 2

KMO Measure of Sampling Adequacy

 The Kaiser- Meyer-Olkin statistics is a sampling adequacy measure that

assesses the strength of the partial correlation of each variable in its assigned factors

(Cerny & Kaiser, 1977). This test measure has a value range of 0 to 1; KMO scores

closer to 1 are regarded as favorable, indicating that the factor analysis will benefit these

variables. This is common when the majority of the zero-order correlations are positive.

While KMO values less than 0.5 are negative correlations that require remedial or

deleting offending variables. After processing the data using Jamovi software, the

proponents obtained a value of 0.717, suggesting that the data is good and acceptable

based on KMO interpretation because it has a significant partial correlation.

Table 3

Bartlett’s Test of Sphericity

Using Bartlett's Test of Sphericity will give a comparison from the observed

correlation matrix to the identity matrix. This will determine whether there is any

redundancy between the variables in order to reduce the number of components. In

terms of checking the null hypothesis, it states that the data come from an uncorrelated

MVN distribution. If the test's p-value is less than the chosen significance level of 0.05,

the null hypothesis can be rejected, and the dataset is suitable for a data reduction

technique (Wicklin, 2022). An identity matrix is a matrix in which all of the values

along the diagonal are 1, and all of the other values are 0. In this case, the proponents

were able to have a significant value of less than 0.001 (<.001). As a result, the

proponents rejected the null hypothesis and are suitable for data reduction as its

significance level is less than 0.05.

Table 4

Eigenvalues

Table 5

Scree Plot
 Table 3 and Figure 4 represent the eigenvalues of each component. From the

data acquired, the total number of components with an eigenvalue of 1 or greater than

1 is 10. The eigenvalue of each component describes the significance of a factor. As

seen in the table, component 1 has the greatest eigenvalue of 5.816, indicating that the

factor with the highest eigenvalue is considered the most significant.

Table 6

Pattern Matrix
 When it comes to the component loadings, items that have a factor loading of

greater than 0.30 load collectively and consequently result measure the same

component. Following the test developers' analysis of the provided data, every item

distributed among the 10 components has a factor loading of greater than 0.30,

indicating a strong association between the test items and the components. It should be

noted that all of the items in the scale's two subconstructs measure the same component

in this context.
VI. References

August, K.J., Rook, K.S. (2013). Social Relationships. In: Gellman, M.D.,

Turner, J.R. (eds) Encyclopedia of Behavioral Medicine. Springer,

New York, NY. https://doi.org/10.1007/978-1-4419-1005-9_59

Aristovnik, A., Keržič, D., Ravšelj, D., Tomaževič, N., & Umek, L. (2020).

Impacts of the COVID-19 pandemic on the life of higher education

students: A global perspective. Sustainability, 12(20), 8438.

https://doi.org/10.3390/su12208438

Bhandari, P. (2022, December 2). Construct validity: Definition, Types, &

Examples. Scribbr. Retrieved December 14, 2022, from

https://www.scribbr.com/methodology/construct-validity/

Collins, S. (2008) Statutory social workers: stress, job satisfaction, coping,

social support and individual differences. British Journal of Social Work

38 (6), 1173–1193.

Conway AM, McDonough SC. Emotional resilience in early childhood:

Developmental antecedents and relations to behavior problems. Ann N

Y Acad Sci. 2006;1094(1):272–77.

Davidson RJ. Affective style, psychopathology, and resilience: Brain

mechanisms and plasticity. Am Psychol. 2000;55(11):1196–214.

Gross, J. J. (2007). Handbook of emotion regulation. New York, NY: Guilford

Press.
Haggerty, R., Sherrod, L., Garmezy, N. and Rutter, M. (1996) Stress, risk, and

resilience in children and adolescents. Cambridge: Cambridge

University Press.

Hurley, K., & Young, A. (2022, July 14). What is resilience? definition, types,

building resiliency, benefits, and resources. EverydayHealth.com.

https://www.everydayhealth.com/wellness/resilience/

Joachim Stöber, Dorothea E. Dette & Jochen Musch (2002) Comparing

Continuous and Dichotomous Scoring of the Balanced Inventory of

Desirable Responding, Journal of Personality Assessment, 78:2, 370-

389, DOI: 10.1207/S15327752JPA7802_10

Krishnan, M. T. (2021, May 7). Understanding Correlations and Correlation

Matrix. AI, Computer Vision, and Mathematics. Retrieved December

14, 2022, from https://muthu.co/understanding-correlations-and-

correlation-matrix/

McDonald, G., Jackson, D., Wilkes, J. and Vickers, M. (2012) A work-based

educational intervention to support the development of personal

resilience in nurses and midwives. Nurse Education Today 32, 378–384.

Morrison, T. (2007) Emotional intelligence, emotion, and social work: context,

characteristics, complications, and contribution. British Journal of

Social Work 37, 245–263.

Stephens, T. (2013) Nursing student resilience: a concept clarification. Nursing

Forum 48 (2), 125–133.

Tavakol, M., & Wetzel, A. (2020). Factor analysis: A means for theory and

instrument development to support construct validity. International

Journal of Medical Education, 11, 245–247.

https://doi.org/10.5116/ijme.5f96.0f4a

Wicklin, R (2022) On Bartlett’s sphericity test for correlation. SAS Blogs

https://blogs.sas.com/content/iml/2022/04/27/bartletts-sphericity-

test.html#:~:text=More%20formally%2C%20Bartlett's%20sphericity%

20test,population%20are%20MVN%20and%20uncorrelated.

Woodward, Kathryn. (2018). Concepts of identity and difference.

https://doi.org/10.4324/9781315668505-34

Zhang M, Lu J. [The development of adolescents’ emotional resilience

questionnaire]. Psychol Sci. 2010;33(1):24–27.