Gem case study

Causation vs Correlation and the Effects of Bias on Research
Questions:
1. What are the independent and dependent variables in this study?
Dependent variable: entrepreneurship activities
Independent variable: activities (promotion, facilitation, long-term commitment to secondary education, developing the capacity of a society to accommodate the higher levels of income disparity, and creating a culture that validates and promotes entrepreneurship throughout society)
2. What are some of the intervening, extraneous, and moderating variables that the study attempted to control with its 10-nation design?
· Innovation
· Culture
· Economy (GDP)
· Geography and politics (geopolitics)
· Proximity to needed supply chains
· Available resources
· Government subsidies
3. Can you do a causal study without controlling intervening, extraneous, and moderating variables?
Extraneous variables should be controlled when possible but it is not always possible to control for them all as they are extraneous. Intervening variables are only hypothetical and are unable to be seen in a study, which is why they are hypothetical. One cannot determine or control their influence. Moderating variables can be controlled to some extent as they are identifiable and may impact the relationship between the independent and the dependent variable (Bauman, Sallis, Dzewaltowski & Owen, 2002; MacKinnon, 2011).
4. What is the impact on study results of using national experts (key informants) to identify and weigh entrepreneurial framework conditions?
The impact on study results of using national experts to identify and weigh entrepreneurial framework conditions is that it introduces bias into the study. It provides a way to measure and entrepreneurial stimulation, but the measures all come from the ways in which the key informants see and judge entrepreneurship. These views may be different among other populations. Experts are not always clinically sound or even 100% objective in their views (Commons, Miller & Gutheil, 2004). One can also have selection bias in terms of what informants to accept as “experts” for the study’s purposes (Pal, Harper & Konstan, 2012).
Bias is a natural condition in man, which is why it is so hard to root it out in a study. As Scripture notes, only Christ is free of all bias in terms of how He looks at people and the world, because He is looking at things from the standpoint of God—and since He is God, He knows what the right way to see things actually is. Two quotes from Scripture that illustrate this point are: “They came and said to Him, ‘Teacher, we know that You are truthful and have no personal bias toward anyone; for You are not influenced by outward appearances or social status, but in truth You teach the way of God. Is it lawful [according to Jewish law and tradition] to pay the poll-tax to [Tiberius] Caesar, or not?’” (Mark 12:14). In this quote, the Pharisees recognize (even though they themselves are acting falsely) the justness of Christ when they point out that He is without bias. They themselves are not without bias, but He is.
And the other quote that emphasizes this idea is: “I solemnly charge you in the presence of God and of Christ Jesus and of His chosen angels, to maintain these principles without bias, doing nothing in a spirit of partiality” (1 Timothy 5:21). Here, it is stated that a Christian has to live according to Christian principles without bias; in other words, the Christian perspective is the way to see all things. Of course, those who are not Christian will point out that this itself is a bias, but as is stated above, there is no way to avoid some kind of bias—so the best thing a study can do is to point out its own bias so that the reader can see it clearly.

5. Can you do a causal study when much of the primary data collected is descriptive opinion and ordinal or interval data?
A causal study can be done when the data is descriptive opinion and ordinal or interval data. Causal studies involve identifying a relationship and then highlighting the nature of that relationship by showing how an independent variable impacts a dependent variable. There is no one right way to do such a study. However, some might argue that descriptive research only indicates correlation—not causation (Curtis, Comiskey & Dempsey, 2016).
Indeed, the debate between causation and correlation is an old one. Can any study genuinely be causal if it cannot control for all types of variables? The gold-star standard of studies in which the sample is a randomized sample and the variables are all controlled would help to show causation best, but so few studies are actually done in this manner. Furthermore, even then it is not 100% guaranteed that all variables have been controlled; thus, one is always more or less highlighting correlation and inferring causation because of the consistency of the relationship between certain variables. Studies are done in order to understand the relationship between variables. Some researchers might hesitation out of an abundance of caution to say that there is a causal relationship between variables based on the type of primary data collected; other researchers will have no such problem because they trust their inferences and believe they are logically deducing a conclusion from the given data.
A study may use descriptive opinion to help show correlation—but most researchers will not accept causation without some statistical analysis. Others see anecdotal evidence as strongly supportive of causation if there is consistency across a wide sample. Still other researchers will want quantitative data. They will want intervals or ordinal data that shows a natural and ordered category of outcomes. Again, there is no one way to pursue a causal relationship in research. It is really up to the researcher and the type of study that the researcher is seeking to do. If the subject of the study would be better approached from a qualitative method, then that is what should be used. If it is better approached quantitatively then that is what should be used. There are all types of ways to approach research for this reason.
References
Bauman, A. E., Sallis, J. F., Dzewaltowski, D. A., & Owen, N. (2002). Toward a better understanding of the influences on physical activity: the role of determinants, correlates, causal variables, mediators, moderators, and confounders. American journal of preventive medicine, 23(2), 5-14.
Commons, M. L., Miller, P. M., & Gutheil, T. G. (2004). Expert witness perceptions of bias in experts. Journal of the American Academy of Psychiatry and the Law Online, 32(1), 70-75.
Curtis, E. A., Comiskey, C., & Dempsey, O. (2016). Importance and use of correlational research. Nurse researcher, 23(6).
MacKinnon, D. P. (2011). Integrating mediators and moderators in research design. Research on social work practice, 21(6), 675-681.
Pal, A., Harper, F. M., & Konstan, J. A. (2012). Exploring question selection bias to identify experts and potential experts in community question answering. ACM Transactions on Information Systems (TOIS), 30(2), 1-28.