A Gem of a Study Discussion
Department of Business, Liberty University
BUSI600: Business Research Methods
Module 2: Week 2
1. What are the independent and dependent variables in this study?
In the CaseBuilder “A Gem of a Study” assignment, researchers from varying professional
sources, including the Kauffman Center for Entrepreneurial Leadership and the London Business
School, sought to determine what factors promote entrepreneurship and, consequently,
perpetuate job, economic, and GDP growth (Schindler, 2019, p. 1). To do so, researchers
designed a study to detect potential causal relationships. As stated by researcher, Joseph
Maxwell, “adequate causal explanations in the social sciences depend on the in-depth
understanding of meanings, contexts, and processes that qualitative research can provide”
(Maxwell, 2012, p. 655). Thus, for their dependent variable, researchers selected to research
economic growth, well-being, and business capabilities/dynamics to gain more in-depth
understanding. Such business dynamics include the creation/loss of new jobs and job growth.
Conversely, they chose as their independent variable socio-cultural and political factors which
may potentially impact entrepreneurial opportunities. This includes constructs listed in the
Exhibit C-Gem 1-1 GEM Conceptual Model such as general national framework conditions,
entrepreneurial opportunities, entrepreneurial framework conditions, entrepreneurial capacity,
and their associated concepts and constructs (Schindler, 2019, p. 2). Each underlying concept,
whether financial markets, perception, skills, or internal market openness, were measured
through survey research, program building, non-standardized data collection, hour-long personal
interviews, or questionnaires.
2. What are some of the intervening, extraneous, and moderating variables that the study
attempted to control with its 10-nation design?
In its attempt to prognosticate economic growth, this study has several intervening,
extraneous, and moderating variables which affect its 10-nation design (Schindler, 2019, p. 2).
As an independent variable which may bias/influence factors a researcher is hoping to examine,
extraneous variables can often “indicate there is a causal relationship when none exists,” causing
researchers concern (Flannelly, L., Flannelly, K, & Jankowski, 2014, p. 165). In this study,
researchers sought to account for these variables by working to measure how they affect
associated dependent variables, consequently grouping test subjects by select features (age,
cultural region, etc.), and including these features in the resulting statistical analysis to “control
for variation in the levels of the variable” (Flannelly, L., Flannelly, K, & Jankowski, 2014, p.
166).
According to the examples of extraneous variables provided by Flannelly and colleagues, the
extraneous variables controlled for in this study include age, cultural background, differing
nationalities, and experience. Those not controlled for could include whether participants have
prior entrepreneurship background, their educational level, their creative ability, whether they
are/are not immigrants, and economic situation (Flannelly, L., Flannelly, K, & Jankowski, 2014,
p. 166). Similarly, moderating variables included within the study can help measure causal
relationships between the independent and dependent variables such as attitudes,
activity/engagement level associated with age, and cultural knowledge associated with country of
origin. Those not controlled for could include inclination and interest associated with quality of
education, economic stability, and cultural backgrounds (Flannelly, L., Flannelly, K, &
Jankowski, 2014, p. 167).
Intervening variables which link and influence dependent and independent variables, include
occupation of the research participants, considering education level/income level have a
proposed association with increased entrepreneurial activity (Schindler, 2019, p. 1). Though not
directly causal in nature, the intervening variable exposes how one variable can influence another
(Flannelly, L., Flannelly, K, & Jankowski, 2014, p. 167). Thus, this 10-nation study design seeks
to control for this disparity by randomly interviewing adults with differing educational
backgrounds from varying countries and interviewing professional experts in the field
(Schindler, 2019, p. 1). Overall, there is no doubt researchers did well in attempting to control for
extraneous, mediating, and intervening variables; yet, it must be determined that there is no way
to completely eliminate the risk for error. Overall, researchers discovered a positive correlation
between perceived opportunity, capacity, and motivation and increased entrepreneurship, GDP,
and employment status; yet, further research could be conducted to better control for variables,
providing greater evidence to establish causal relationships (Schindler, 2019, p. 4).
3. Can you do a causal study without controlling intervening, extraneous, and moderating
variables?
According to Chambliss and colleagues, a causal study can be conducted without controlling
for intervening, extraneous, and moderating variables (Chambliss & Schutt, 2003, p. 108).
However, it should be noted that the validity and reliability of the study may be less accurate,
and that, to achieve confidence in results, one must meet the three requirements of: empirical
association, nonspuriousness, and independent variable temporal priority. Within this study, it is
clear that the first criterion, association, is met as several positive correlations are identified with
entrepreneurship, including the perception of opportunity, motivation, and capacity. The
inclusion of the second criterion, time order/temporal priority, within this study is questionable
as this condition would require manipulation of the independent variable to come into being
before changes were observed with the dependent variable. In this study, it cannot definitively be
stated whether perception of opportunity, capacity, or motivation came prior entrepreneurship
activity or were a result of positive results from such activity. This supports the idea that a
longitudinal study would be most effective, including a study conducted on the same test group
pre and post business start-up. Such research would help fulfill the third criterion of
nonspuriousness, otherwise known as “authentic correlations suggesting potential causation”
(Chambliss & Schutt, 2003, p. 109). These three criteria must then be examined through
experimental research to best detect a causal relationship. This would include testing time order
of changes between the independent/dependent variables, testing between a control and
experimental group, and randomly assigning members to the two groups (Chambliss & Schutt,
2003, p. 109).
4. What is the impact on study results of using national experts (key informants) to identify
and weigh entrepreneurial framework conditions?
Utilizing national experts (key informants) to help weigh and identify the basis for an
entrepreneurial framework, researchers requested they complete a standardized 12-page
questionnaire as a means of gleaning indirect information regarding their observational
experiences, including sociocultural, political, and attitudinal influences. Also retrieving data
from personal interviews, researchers were able to collect additional data not retrievable through
a questionnaire. In using key informants, this study does well in helping researchers “study
collective behavior” -- the purpose for which such study was intended (Hughes & Preski, 1997,
p. 82). Obtaining this first-hand information from those knowledgeable in the field, helps shed
light on entrepreneurship perception and action across cultures. However, as the number of
experts interviewed varied within each country, ranging anywhere from 4 to 39, these results
may not totally be indicative/representative of their native country’s views (Hughes & Preski,
1996, p. 82). In addition, there is no information provided regarding potential biases key
informants could bring to the research or even by what guidelines researchers considered them
experts. As research exposes that “several types of bias are of particular concern when using key
informants for research,” this further brings about questions as to the whether the impact of using
them increases/decreases the validity of the results (Hughes & Preski, 1996, p. 82). Overall, this
information provided data for comparing entrepreneurship perception in differing countries
though validity is questionable. Thus, to improve this study in future, experts could be chosen
based on preselected, established guidelines and a similar number within each country could be
selected to help lessen potential cultural biases.
Can you do a causal study when much of the primary data collected is descriptive opinion and
ordinal or interval data?
In assessing this study and its components, it could be determined that this study is a
descriptive study with potential for future causal research, and is not, in and of itself, a causal
study. According to research conducted by Lynn, such descriptive study should be supported
with additional information before assuming causation as he states, “popular opinion and
procedures are not necessarily reliable guides to best practices” (Lynn, 2002, para. 2). Instead, he
encourages researchers to “submit manuscripts that report the results of correlational studies and
experiments designed to identify the causes of outcomes,” promoting the idea that
experimentation is needed to determine true causation (Lynn, 2002, para. 7). Though GEM
researchers do their due diligence to account for extraneous, intervening, and moderating
variables, validity and reliability must be questioned for these variables cannot be fully
controlled for. Additionally, as the majority of the primary data collected is descriptive opinion
and ordinal/interval data, validity and reliability associated with implying causation when only
correlations are specified must come into question (Lynn, 2002, para. 1). Yet, as this study
presents qualitative data through statistics, it does provide semi-measurable elements which
could be used as a basis for causal research. Thus, it could be determined that this study should
serve as a segue for further studies.
In exposing positive correlations between select independent and dependent variables, this
study fails to provide solidified research to imply causation (Schindler, 2019, p. 4). To rectify
this, these correlations should be studied over time rather than on a short-term basis. In addition,
increased representation of countries could be included, expanding from the 10-country design
(Chambliss & Schutt, 2003, p. 108). With lesser represented countries, including Africa and
Mexico, not included, it may be argued that these cultures are not accurately represented,
skewing results. Further affecting the validity of results, researchers attempt to measure multiple
indices at once verses manipulating and studying one variable at a time. To enhance this process
and lead to further evidence for causation, it would be beneficial to manipulate one variable at a
time and research its associated effects (Chambliss & Schutt, 2003, p. 108). This search for the
truth through planned, effective research should be taken seriously for it is not only supported
through science, but the Bible. As David states in Proverbs 25:2, “It is the glory of God to
conceal a thing: but the honour of kings is to search out a matter” (King James Bible,
1769/2017). Thus, for the researcher, one must not cut corners to imply causation from
correlation, but must use such information as a platform to search for further knowledge.
References
Chambliss, D. & Schutt, R. (2003). Causation and Experimental Design. In Jeff Lasser (Eds.).
Making sense of the social world (6th ed., pp. 106-135). Thousand Oaks, California: SAGE
Publications.
Flannelly, L., Flannelly, K., & Jankowski, K. (2014). Independent, dependent, and other
variables in healthcare chaplaincy research. Journal of Health Care Chaplaincy, 20(4), 161-170.
Hughes & Preski. (1997). Using key informant methods in organizational survey research:
Assessing for informant bias. Research in Nursing & Health, 20, p. 81-92.
King James Bible. (2017). King James Bible Online.
https://kingjamesbibleonline.org/ (Original work published 1769)
Lynn, Mike. (2002). The industry needs less descriptive and more causal research. Cornell Hotel
and Restaurant Administration Quarterly, 43(2).
Maxwell, J. (2012). The importance of qualitative research for causal explanation in education.
Qualitative Inquiry, 18(8), 655-661.
Schindler, P. (2019). Business research methods (13th ed.). McGraw-Hill.
Powered by TCPDF (www.tcpdf.org)
