Description
Include all JASP output for reliability and validity in the Appendix.
Include a .csv file with your cleaned survey data.
With your dissertation topic in mind, design a survey that could be used to gather data to answer a possible research question. The survey should include ONLY TWO subscales and include 8 Likert scale questions. Consider the length of the survey, the number of questions, and the scale that could be used. The assignment requirements include:
Design a survey with 8 Likert scale questions (10 at most) and TWO subscales. Each subscale needs at least four questions.
Have at least 10 individuals complete the survey. The more people that complete your survey, the better. Also, feel free to create an online survey for easier distribution.
Using JASP and the collected survey data, determine your survey instrument’s validity (Exploratory Factor Analysis) and reliability (Cronbach’s Alpha).
In your write-up, include the following:
Research topic and name of the survey you created
Explain the two survey subscales and what they should be measuring.
Provide a list of all survey questions and answer options (in Appendix)
Which questions go with which survey subscale.
Survey Methods
How did you collect your data? Online, Face-to-face?
Who did you collect your data from?
How many people did you collect data from?
Cronbach Alpha reliability results in APA format. What would you do to improve the reliability? Try to get the highest Cronbach’s Alpha value you can.
Exploratory Factor Analysis (EFA) validity results in APA format. What would you do to improve the validity?
What are your thoughts on the survey/process?
Include all JASP output for reliability and validity in the Appendix.
Include a .csv file with your cleaned survey data.
Have at least one citation (peer-reviewed and/or textbook) to support your discussion.
NOTE: For this assignment, you are doing an EXPLORATORY Factor Analysis, not a Confirmatory Factor Analysis.
Unformatted Attachment Preview
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
Expatriate Teachers’ Sense of Cultural Belonging
James Riley
University of the Cumberlands
DSRT 834: Advanced Statistics
Doctor Kona Jones
June 18, 2023
1
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
2
Expatriate Teachers’ Sense of Cultural Belonging
Introduction
The research being conducted for the approved dissertation will investigate the
relationship between transformational leadership and cultural belonging. This survey instrument,
“Expatriate Teachers’ Sense of Belonging,” endeavors to measure expatriate teachers’ sense of
cultural belonging. This survey was developed with two subscales, group membership (GP) and
national privilege (NP). The four questions in the group membership subscale should measure
the extent to which expatriate teachers feel that they are part of and identify with the target
culture. The initial four questions in the national privilege scale endeavor to measure the extent
to which expatriate teachers feel they are treated the same and enjoy the same privileges as local
teachers. The complete eight question survey is listed in appendix A. The survey was
electronically distributed through email, social media, and WhatsApp to friends and former
colleagues. The survey was sent to fourteen people, and ten people completed the survey (71.4%
response rate).
Reliability
Cronbach’s alpha coefficient was checked for the “Group Membership” subscale. For the
“Group Membership” subscale, Cronbach’s α = 0.839. While there is no universally accepted
minimum reliability level, a Cronbach’s alpha greater than 0.70 is considered acceptable in many
fields of research (Bonett & Wright, 2014). Therefore, the “Group Membership” subscale was
found to have high internal consistency based on α = 0.839. JASP did indicate that the internal
consistency could be improved by removing question GM7, which would raise Cronbach’s α to
0.857. However, removing any additional questions would lower Cronbach’s alpha score, and
thus lower the measure of internal consistency.
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
3
Cronbach’s alpha coefficient was also checked for the “National Privilege” subscale. For
the “National Privilege” subscale, Cronbach’s α = 0.6. Since a Cronbach’s alpha greater than
0.70 is considered acceptable in many fields of research (Bonett & Wright, 2014), the “National
Privilege” subscale was not found to have high internal consistency. However, JASP did indicate
that the internal consistency could be improved by removing question NP4, which would raise
Cronbach’s α to 0.845. However, removing any additional questions would lower Cronbach’s
alpha score, and thus lower the measure of internal consistency.
Validity
An exploratory (EFA) was done to check the validity of the “Teachers’ Sense of Cultural
Belonging” survey subscales. The initial model with all eight questions was not significant ( p =
0.579). Further, Bartlett’s test showed that these questions were not a good fit for factor analysis
(p = 0.07). A parallel analysis based on factor analysis in JASP indicated one factor for this
survey. The survey was developed with two subscales (GM and NP), and this was confirmed
through a scree plot. Based on this scree plot, a manual analysis with two factors was performed
in JASP. All questions were loading on the appropriate factor, with the exception of NP4, which
did not load on either factor. As a result, question NP4 was removed, and a final test was run.
Bartlett’s test revealed that these questions (without NP4) were a good fit for factor analysis (p =
0.015). This two-factor model was not found to be significant X2 (8) = 9.038, p = 0.339.
As shown in the factor loading table below, each question loaded on the appropriate
subscale. Questions NP2 and NP6 had low scores indicating that they were not a good fit with
that factor. To increase the validity of the survey, these two questions could be removed.
However, that would reduce the survey only one question on national privilege (NP). Further,
this dramatic decrease in the number of questions could impact the reliability of the survey.
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
4
Therefore the questions and their intended purpose should be reviewed to see how they can be
adjusted to best measure national privilege.
Discussion
I am pleasantly surprised by the results of this survey and the data. I am disappointed that
the Chi squared analysis did not come back as significant. However, I was happy that all the
questions loaded based on their assigned factor. Further, I only needed to drop one question. As I
was designing the survey, I had a bad feeling about that question (NP4). I see the value in going
back to previously developed and tested survey instruments. For my own study, nothing that has
been previously created is a perfect fit. However, I can adapt certain measures from pre-existing
surveys and develop research instruments that fit my unique needs.
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
5
References
Bonett, D. G., & Wright, T. A. (2014). Cronbach’s alpha reliability: Interval estimation,
hypothesis testing, and sample size planning. Journal of Organizational Behavior, 36(1),
3–15. https://doi.org/10.1002/job.1960
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
6
Appendix A
Expatriate Teachers’ Sense of Cultural Belonging Survey Instrument
Each of the following each questions was administered in a 5-point Likert scale format.
1 = Strongly Disagree
2= Disagree
3 = Neutral
4= Agree
5 = Strongly Agree
1. I feel at home working in my school. (group membership)
2. As an expatriate teacher, my input is highly valued. (national privilege)
3. In my school, I am part of a team. (group membership)
4. I experience more restrictions as a foreign teacher. (national privilege)
5. Teaching at this particular school is important to me. (group membership)
6. The opinions of local teachers are more highly valued than my own. (national
privilege)
7. I expect to be part of this school for a long time. (group membership)
8. I am treated the same as local teachers. (national privilege)
The following questions address the subscale of “group membership:”
Question 1, 3, 5, and 7.
The following questions address the subscale of “national privilege:”
Question 2, 4, 6, and 8.
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
Appendix B
JASP Output for Reliability
This initial test for group membership showed that GM7 could be removed to enhance
reliability.
After removing GM7, no further adjustments would improve survey validity.
7
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
This initial test for national privilege showed that NP4 could be removed to enhance
reliability.
After removing NP4, no further adjustments would improve survey validity.
8
EXPATRIATE TEACHERS’ SENSE OF CULTURAL BELONGING
9
Exploring the Impact of Promoting Awareness on Purchasing Behavior:
Promoting Awareness and Purchasing Behavior Survey Study
Yujia Bai
University of the Cumberlands
DSRT-834-A01 Advanced Statistical Application
Dr. Kona Jones
06/18/2023
Introduction
Marketers have never made less effort to find a way to drive consumer demand and
change purchasing behavior. Among many marketing strategies, offering promotion seems the
most efficient way to change customer behavior. The price and quality of the products customers
perceive are essential factors in purchase decisions (Haghighatnia et al., 2018). But will the
promotion work equally for the customers with different awareness levels of promotion
activities? Thus, this survey explores the impact of promoting awareness of purchasing behavior.
Methods
To explore the relationship between promoting awareness and purchasing behavior, this
research uses a questionnaire to collect data on people’s promoting awareness and purchase
behavior based on promotions. An online survey powered by SurveyMonkey was used, and
some Excel questionnaires were sent via online message platforms. As a result, there are total 16
valid responses collected. The participants are a young-age group ranging from 20 to 35 years
old and mixed gender.
There are two subscales (A – Promoting Awareness and B – Purchasing Behavior) with
four questions in each subscale. For each question, the participant needs to choose from multiple
options based on a 5-point Likert scale (1 – Strongly disagree, 2 – Disagree, 3 – Neither agree nor
disagree, 4 – Agree, 5 – Strongly agree).
Survey questions (SurveyMonkey link):
•
Subscale 1: Promotion Awareness (A)
A1. I’m always aware of promotion information.
A2. I usually subscribe and review the promotions from store/retailer.
1
A3. I think knowing potential promotions before purchasing is important.
A4. I search for potential discounts before placing an order.
•
Subscale 2: Purchasing Behavior (B)
B1. Whether there is a discount or not will determine my purchasing decisions.
B2. When there’s a promotion, I tend to purchase more than I need.
B3. I tend to purchase unnecessary things because of promotions.
B4. Stuff in discount makes me feel it is lower quality.
Results
Reliability test
Cronbach’s alpha coefficient was checked for the “Purchase Behavior and Promoting
Awareness Survey” survey’s “Promotion Awareness (A)” and “Purchasing Behavior (B)”
subscales. For “Promotion Awareness (A)” subscale (A1, A2, A3 and A4), Cronbach’s α = .729
(see Appendix 1). For “Purchasing Behavior (B)” subscale (B1, B2, B3 and B4), Cronbach’s α
= .656 (see Appendix 2). A Cronbach’s alpha greater than .70 is considered acceptable, so the
“Promotion Awareness (A)” subscale of “Purchase Behavior and Promoting Awareness Survey”
survey was found to have high internal consistency while “Purchasing Behavior (B)” subscale
was not.
Even though “Promotion Awareness (A)” subscale has high internal consistency, JASP
indicated that Cronbach’s alpha could be improved to as highest as .917 if A1 and A2 removed
(see Appendix 3). Besides, to improve the internal consistency of “Purchasing Behavior (B)”
subscales, JASP indicated that if B1 and B4 were removed, the internal consistency would
2
improve to α = .832 (see Appendix 4). Since the improved Cronbach’s alpha of “Purchasing
Behavior (B)” subscale is greater than .70, this subscale will have high internal consistency if B1
and B4 removed.
Validity test
An exploratory factor analysis (EFA) was done to check the validity of the “Purchase
Behavior and Promoting Awareness Survey” survey subscales. The model was found not to be
significant, χ2(28) = 55.957, p = 0.527, indicating the model is not significant. Bartlett’s test
showed a significant p-value (p = 0.001), indicating that the survey questions were a good fit for
factor analysis. The survey was developed with two subscales, but the EFA indicated that A2
cannot fit into any subscales while other factors could fit into 2 subscales. See below for the
factor loadings of each subscale.
As shown on the Factor Loading table, some of the questions had low scores, indicating
that they weren’t a good fit with that factor. These questions included B1, B4, and A2. To
increase the validity of the survey, it is recommended that these three questions are removed. In
addition, in looking at the loading results, Factor 2 is loading with the B subscale, indicating a
good fit. Factor 1 has questions loading from both A and B subscales. This seems to indicate that
the questions are not different enough to warrant separate subscales. It is important to review the
3
questions and intended purpose of the questions to see if they need adjusted. Another potential
issue is with the sample size for the survey, Gorsuch (1983) recommended 100 response
minimum sample size for Factor Analysis. This survey only included 16 participants, which is
well under the recommended minimum number.
Conclusion
In summary, this survey is reliable but needs a better validity result. Moreover, the large
p-value of chi-square indicates the model is not significant, which means the research question
needs to be re-evaluated since the observed result differs from the expected results. Besides, the
question A2 “I usually subscribe and review the promotions from store/retailer”, B1 “Whether
there is a discount or not will determine my purchasing decisions”, and B4 “Stuff in discount
makes me feel it is lower quality” impact both reliability and validity results negatively. Thus,
those three questions need to be considered to be revised or removed from the current two
subscales. However, as mentioned above, the insufficient sample size might be the root cause of
all problems, and a larger sample size needs to be collected to make the final decision.
4
Reference
Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Lawrence Erlbaum.
Haghighatnia, Abdolvand, N., & Rajaee Harandi, S. (2018). Evaluating discounts as a dimension
of customer behavior analysis. Journal of Marketing Communications, 24(4), 321–336.
https://doi.org/10.1080/13527266.2017.1410210
5
Appendix
Appendix 1: JASP reliability results of subscale A
Appendix 2: JASP reliability results of subscale B
6
Appendix 3: JASP reliability results of subscale A (A1, A2 removed)
Appendix 4: JASP reliability results of subscale B (B1, B4 removed)
Appendix 5: JASP EFA validity results
7
Purchase answer to see full
attachment
