Uploaded by Kevin Wanjala Nyongesa

Survey Design

advertisement
SURVEY DESIGN
1
Definition of the Study
Does high self-esteem bring about the better academic performance of undergraduate
university students? This survey sought to establish how high self-esteem affects the academic
performance of undergraduate university students and effectively examine the relationship that
exists between high self-esteem and the academic achievement of undergraduate university
students. Self-esteem is identified as the independent variable in the study, and academic
performance as the dependent variable.
In psychology, self-esteem is conceptually defined as the personalized attitude towards
self-worthiness or the individual’s desirability. In other words, it is the degree of self-appreciation
and likeness which includes the inner beliefs, emotional state, and behavior. Operationally, selfesteem can be defined as what the mind believes to be factual and the extent of self-worthiness,
capability, and how valuable an individual is and which is pegged on the effect of environmental
factors. Self-esteem can be measured by how well the mind responds to environmental factors.
The interrelation of self-esteem and the environment can be measured by the following
components; internal locus of control, sense of inclusion, and one’s approval, and the feeling of
proficiency. Self-esteem can be subjected to the nominal level of measurement as it is considered
to either be high or low. Academic performance on the other hand is the degree of achievement of
either short-term or long-term educational objectives. It is the assessment of the student’s
individual or school overall achievement across numerous educational subjects. Academic
performance is well appraised by achievement levels of students, discipline referrals, frequency of
school attendance, rates of graduation, and the level of teacher satisfaction which are considered
to be ordinal variables.
SURVEY DESIGN
2
The study utilized the following research hypothesis to guide the course of the research
and establish the relationship between the variables; ‘High self-esteem does not bring about better
academic performance’.
This is a directional hypothesis that dictates the direction of the
anticipated relationship. This direction gives a rough approximation regarding the relationship
between the two variables; self-esteem and academic performance. However, a zero correlation
exists between the two variables as per the hypothesis that is, no relationship between high selfesteem with better academic attainment of undergraduate university students. A null hypothesis
is significant to this study because it is the basis for argument and seeks to remove the doubt about
the validity of the relationship that exists among the variables.
Methods
The survey design was adopted to draw the link between high self-esteem and the
educational achievement of undergraduate university students. The design was used to
quantitatively describe distinct characteristics of a given population while examining the
relationship that exists among the variables. In this study, the distinct characteristics will be
collected from a sample of the population and the results retracted back to the entire population.
The study will target undergraduate university students as primary participants. When gathering
the sample from the target population, the study will adopt a stratified random sampling technique.
It involves dividing the population into smaller groups. This technique will ensure the
representation of the key factors. 300 undergraduate students within the age of 18 to 25 years will
be chosen as the target population. The stratification of the target population was based on the
cases of extroverts among the students being given the priority. This will be the inclusion criteria
for the stratification of the population. The population was made up of both genders each taking
half making sure that both genders are covered in the survey. The sampling frame in this process
SURVEY DESIGN
3
will be the reported extrovert students within the university. This will be used as a benchmark of
the students with higher self-esteem in the university. To ensure, an equal age distribution in the
sample, three age groups will be created, 18 to 20, 21 to 23, and 24 to 25 years. Equal percentages
from the age groups will be selected arriving at a total sample size of 100 students. Stratified
sampling ensures the precision and inclusiveness of the findings by minimizing the sampling bias.
However, it requires prior comprehension of the aspects of the sampling frame which might not
always be available and sometimes finding the basis of stratification is difficult.
To effectively collect data from the sample population, a structured questionnaire with a
Likert scale of measurement of self-esteem was developed. This measurement scale gauged the
self-esteem by giving the participants assertions depicting the beliefs about the self-esteem being
measured. The Likert Scale has a series of descriptions of opinions over an issue. The participant’s
self-esteem is then a degree of how the respondent agrees or disagrees with the opinion. The
measurement scale is a 5-point scale of how the participants agree or disagree with the suggestion.
This structured questionnaire presented standardized questions to the respondents with no
variations. The questionnaire was suitable for this survey study because of its ability to easily
collect data from a large sample and it provides quantitative data which is easy to collect and
analyze. The developed questionnaire adopted a funnel format of a survey which commences with
general questions about the self-esteem of the respondent and then narrow down to more specific
aspects of self—esteem, hence requiring more details at subsequent levels of the survey. Since this
survey involves examining the respondents’ attitudes, beliefs, and self-esteem at large, a self-report
survey will be adopted. This survey required individual respondents to respond to the questions by
themselves. A self-report survey is suitable because it enables the collection of data about the
respondents’ behaviors which can be regarded as private or unethical when simulated in the
SURVEY DESIGN
4
laboratories. However, due to the private nature of information being sought from the respondents,
some data might be exaggerated thereby skewing the results of the study.
To achieve a successful survey, certain principles have to be adhered to. Concerning this
study, the following rules were followed; defining a clear and measurable objective for the survey,
which in this case was to measure the level of self-esteem among students, focusing on closedended questions which involve the use of pre-populated response choices for the participant to
choose from hence providing comprehensive quantitative data for easy classification and analysis,
keeping the more personal or private questions towards the end, which ensures that the respondent
is bombarded with personal queries at the start which might create discomfort that might lead to
the respondent concealing some of the crucial data, and lastly making every question in the survey
count by making sure that each inquiry contributes and guide the responses of the survey which
are directly related to the objective of the survey. This was accomplished by adopting a funnel
survey format that is suitable for the collection of personal and private information from the
participants and also aimed at creating comfort for the participant before up close questions are
introduced.
Critique
The validity of this survey however was vulnerable to the threats. These are the
inaccuracies and biases in the articulation of the research problem, study design, sampling
procedure, and the collection of data which may result in a systematic variance from the true
findings of the population, and unavailability of the validity in the survey makes the dependent
and confounding variables to influence the behavior of the dependent variable. The risks to the
viability of this survey are; firstly, selection. This is when the respondents in the survey have
divergent attributes such as age, attitude, intelligence among others and the uniqueness might skew
SURVEY DESIGN
5
the results. Secondly, history, that is the occurrence of the events either concurrently or previously
which might amend the findings of the study, and lastly, the attitude of the respondents. This is
such that, the suggestions of the respondent change due to their involvement in the study; also
known as the Hawthorne effect, thus altering the findings of the survey.
Furthermore, the survey was not free of confounding variables. These are the overlooked
intruding factors that are closely related to the independent variable which may ruin the survey or
tend to skew the findings by influencing the responses during data collection. Stress levels among
the students proved to be a confounding factor during the survey and which could influence the
responses by the participants. A respondent under stress would tend to exaggerate the responses
due to the cloud effect resulting in false positive or negative responses. To counter the effects
posed by the confounding factor, the study adopted randomization and restriction strategies during
sampling. Randomization minimized the capability of confounding by coming up with fairly
comparable strata concerning the confounding factors while restriction eliminated the deviation in
the confounding variable by limiting the study to those students who appeared not disturbed by
the subject under study hence they were free to give the required information.
In conclusion, the survey was the most appropriate design to be adopted during the study.
The design was inclusive in the number of variables to be studied, and minimal investment was
required to develop and implement thus making it easier to draw the conclusions. The survey
design was capable to collect data from large samples of the population. Unlike the experimental
design, the survey elicited information about the attitude, beliefs, and behavior that make up the
self-esteem of the respondents. Coupled with the self-report procedure, the survey collected
information first-hand information which was deemed to be personal and might be considered if
simulated in laboratories. The survey however did not fall short of limitations, these included;
SURVEY DESIGN
6
vulnerability to errors and biases which occurred from the unavailability of responses from the
participants or the inaccuracy of the received responses, errors resulting from intentional twisting
of behaviors by the participants to align with the prerequisites of the survey, and the inability of
the respondents to the conditions their behavior thus giving rise to the confounding factors. This
survey would be relevant to the psychologists and academicians to establish the underlying factors
concerning the academic performance of students and drawing up necessary measures aimed at
uplifting self-esteem among students and thus their academic achievement.
Download