The introduction includes a section on scientific relevance. The question you should answer in the paragraph about the scientific relevance is: “What will this new research add to the existing body of literature, and why is that an important addition?”. In other words: what is the innovation in your research?
Prototypical arguments about scientific relevance are:
- You will Discover: there are no data available in previous research about the phenomenon that you study.
- You will Replicate: previous research has concluded that X influences Y. You will check whether the same relationship can be found in another set of observations.
- You will resolve an Anomaly: there are observations that seem to reject existing hypotheses or theories. Your research will clarify what is going on.
- You will solve a Mystery: we do not (fully) know how to explain Y. Your research will add to a piece to the puzzle.
- You will follow Leads: we think we know that X influences Y, but we are not sure. Your research will show to what extent X influences Y.
- You will open a Black Box: we do not how to explain the correlation between X and Y. Your research will show how X influences Y, through which intermediary variables.
- You will use Better Methods: previous research has relied on research designs that are not fully adequate to answer the research questions. Your research will use more refined, sophisticated, and stringent data and methods.
- You will Reopen a Closed Case: previous research has concluded that X influences Y, but there are reasons to doubt this conclusion. Your research will show to what extent X really influences Y.
- You will Generalize: previous research about X and Y was about context A, but you study context B.
Let’s go over each of these types of arguments one by one.
1. Discover
Your research can contribute by charting unknown territories, collecting data on phenomena that previous researchers have talked about but have not actually described. By providing the first empirical analyses of a certain phenomenon, you may provide the ground work for future research. Of course you have to make sure that you are really the first one (see the section on finding literature here).
2. Replicate
If previous research came to a surprising conclusion on the relationship between two variables, it would be good to check whether the same conclusion emerges from a similar analysis of other data, including the same variables. Don’t be surprised if your replication generates results that differ from previous research. Due to publication bias, p-hacking, and other forms of research misconduct and flaws in the process of academic research, replications often ‘fail’ (Simmons, Nelson & Simonsohn, 2011; Open Science Collaboration, 2015). Also if previous research has yielded mixed results it is illuminating to replicate. Perhaps your research can show why previous research has yielded mixed results. Even if that is not the case, it is interesting to have an additional piece of evidence on the relationship.
3. Resolve an Anomaly
A good starting point for a scientific discovery journey is an anomaly: an observation that cannot be explained by prevalent theories in your field because it the observation runs counter to predictions from these theories. In research on volunteering for instance it has frequently been observed that persons who earn higher incomes are more likely to volunteer than persons with lower levels of income. This runs counter to the opportunity cost theory of volunteering (Menchik & Weisbrod, 1987; Wilson, 2000; Carlin, 2001). Persons who earn higher incomes lose more by providing their labor without monetary compensation and should therefore be less likely to volunteer, everything else held constant. The question why people with a higher level of income volunteer more poses a puzzle for the opportunity cost theory.
4. Solve a Mystery
Perhaps an even more compelling starting point for scientific discovery is a mystery: a phenomenon that we do not understand and cannot explain. The discovery of a phenomenon may give rise to a mystery if no theory is able to explain its occurrence and origins. The mystery bears resemblance to the anomaly: it is also a puzzle. The difference between an Anomaly and a Mystery is that we have no theoretical basis from which we perceive the phenomenon as impossible or unlikely. The best mysteries are those for which we have clue where to start searching for an answer, and no evident starting point of explanation. Total mysteries are rare. In the more likely case there is a theory (or a set of theories) that does not fully explain why a certain phenomenon occurs. Your job is to sort out what additional explanations provide a more complete account of the phenomenon.
5. Follow Leads
In some respects, science is like solving a murder case. You follow leads to find the person guilty of the murder. The best way to find the killer is to follow multiple leads, and to stop searching when a lead does not seem to be promising any more. Following only one lead may lead to tunnel vision in which all evidence is interpreted as a confirmation of the initial suspicion.
The analogy with science is the terrible trend towards more publications of positive results: a growing proportion of papers in (international, peer-reviewed, and high ranking) academic journals is reporting confirmations of hypotheses rather than rejections (Fanelli, 2012). Support for a certain – often new – hypothesis adds less to our body of knowledge than rejection of an old hypothesis.
In other respects, science is like criminal justice as well. In the legal system, convicting an innocent murder suspect is viewed as a more severe disadvantage than setting free a killer due to lack of evidence. Avoiding false negatives (type II errors) is viewed as more important than avoiding false positives (type I errors).
Table 1. Errors in hypothesis testing
| Hypothesis is not true | Hypothesis is true | |
| Hypothesis accepted | Type I error False positive Convicting an innocent suspect | No error True positive Convicting a killer |
| Hypothesis not accepted | No error True negative Setting free an innocent suspect | Type II error False negative Setting free a killer |
6. Open a Black Box
When previous research concluded that two variables are related, the question is how the relationship can be explained. Often you can find arguments in previous research about why the relationship exists without an explicit test of this argument. You can open the black box by formulating and testing hypotheses about the variables that mediate the relationship. The strongest contribution you can make in such a situation is to formulate multiple hypotheses about mediating variables. Figure 4 below shows a model of the relationship between religious affiliation and volunteering, answering the research question why protestants are more likely to volunteer than non-religious persons. The model includes two different mediating variables: solicitation and altruistic values.
7. Use Better Methods
Innovations can also be in the methodology you use. If you use better methods to evaluate a hypothesis that has been proposed earlier than previous studies, then you are innovative. For instance, if we do not know whether the correlation between X and Y reflects a causal influence of X on Y and your research uses longitudinal or experimental methods that get at the direction of causality in the relationship, it is innovative.
8. Reopen a Closed Case
Sometimes the literature provides a very clear answer to a research question, but you have reasons to believe that the consensus is ill-founded. New data or the use of better methods and research designs may cast doubt on commonly held beliefs. In these cases, your research provides new insights.
9. Generalize to New Cases
Previous research has examined the causes you are interested and their consequences in certain times, places, and persons or organizations. Your research studies familiar Xs and Ys, but in a different and place, and among different actors.
10. Pitfalls in descriptions of scientific relevance
Examples of phrases that suffer from typical problems in formulations of scientific relevance are:
1. “This phenomenon has never been studied before.” This is not a good argument. It is merely a statement of fact – one that you should check before you write it down, by the way. An example may serve to illustrate why it does not apply in general. We don’t know much about the secret life of firebugs. There is a very good reason why we don’t know much about the secret life of firebugs: it is totally uninteresting!
 |
| Firebugs may look dangerous, but they are harmless animals. They would be interesting if they provide an Anomaly, a Mystery, Leads, or if knowing more about them opens a Black Box. |
2. “This phenomenon has been studied frequently in previous research.” Without additional arguments this is not an adequate justification of the scientific relevance of your research. If there is so much prior research, then what does your research add to it?
3. “This phenomenon has recently attracted a lot of attention from scholars.” Well, why have others found the phenomenon interesting? Do they give sound arguments? Without any additional arguments you may be saying that you are following the mistakes of others.
4. “This phenomenon has generated heated debates in the media.” The fact that people are debating about an issue does not in itself justify attention from scholars. Media attention because people are concerned about an issue is a sign of societal relevance, not scientific relevance. However, it can be very useful to inform the public debate by the facts. The public debate can be ill-informed, misguided by incorrect theoretical assumptions, or methodological pitfalls such as claiming causality on correlational evidence or unjustified generalizations to a broader population. In this case, scientific attention is warranted to inform the public debate.
5. “I am a … myself.” In many cases people study phenomena that they are somehow personally involved with and do so because of that personal involvement but do not report this. Scrutinize the roots of your research questions and try to evaluate in all honesty how this personal commitment may affect your judgment. You need not be an immigrant to study remittances or a high net worth individual to study philanthropy by the wealthy; neither do you need to be a completely impartial observer. It does help, however, to review your own arguments again taking the position of such an impartial observer.
6. “This study contributes to the literature.” Do not overpromise your contribution and what the contribution is exactly. When you plan your research, you may think you can make all sorts of contributions. It is good to start ambitiously, but when you have completed your research, reread your introduction and revise your statements. The contributions you thought you could make may turn out to be less revolutionary. That is fine. A typical outcome of research is that things are more complicated than they seemed at the beginning.
Better Academic Research Writing 