Longitudinal Studies and Dissertations: Your top 5 Mistakes

When you see the word ‘longitudinal study’ or ‘dissertation’ it makes you weep on the inside, right?

It is is an energy sucking and time thirsty endeavor, but once you make it through, it is like reaching the land of Milk and Honey (of journal publications). Factoring in some flexibility is important since outcomes or obstacles we could not have possibly foreseen start to pop up like trolls from under a bridge. Crossing it in a spring is impossible, but having a strategy to avoid or appease some trolls is your best bet.

ERROR ONE: If it Quacks, it Could Be a Theatrical Chicken

An erroneous measurement is not necessarily something that is unreliable, but can arise from behaviors of the subjects or the researchers themselves. Be sure to be aware of these behaviors or keep them in mind. Not everything is strictly just data, if they could be slanted by such things as environment or mood. That data might be good, but the methodology that derived that data may not be.  Measurements, in the way that a researcher may first define them, may be perceived as platonic at first, when they are not as the study continues forward. If people agree with a definition of success, for example, the assumption is that they will not change their definition over time based on certain factors throughout the duration of the study. Being aware of latent variables and how they impact measurement is important. One of the most under written, and quickly rushed through sections of research  is the section describing and justifying the measurement tools. Did you test and re-test these tools? Do they present any errors, and if so, under what conditions? If this is a longitudinal study, be prepared to discuss how your tools will be calibrated and re-calibrated as the implications of error can not only sway your study but can also shape the research of others who base their study off yours! If you can point out the weakness of your study–good for you! That is not a weakness. That is a call to rally other researchers in pooling together brain cells and coming up with a better methodology or measurement tool. Science evolves and it comes in steps. Research is never in one go–and as your literature review will show you, it takes generations.

ERROR TWO: Literature Reviews are not Glossaries with Descriptions

One of the biggest errors is an incomplete literature review. A literature review is not merely writing a summary of the top most cited studies in your field. If you are doing a literature review to justify the means, then the means to completing your study will be filled with holes and unforeseen variables. Why? Because literature reviews are meant to be organic. They lead you to decide what it is you are exactly setting out to study and help you refine your research question, your theoretical framework, and the type of tools or data you strive for. If you are doing a literature review to justify your already scripted research question, then you will only search for justifications rather than discovery. Explore and look for tangents in studies as you research prior published findings. Especially look at discussions, since these often reveal questions the researchers suggest that future researchers pursue. Assess where the researchers did not examine a certain variable, did not consider a theory, or overlooked a relationship or summarized it for the sake of convenience. Sometimes researchers take a completely unrelated theory or philosophy and try it out in their field of study. That is where we see innovation and new considerations take root. And where you find some gaps in the reasoning of other researchers, you will not be critiqued as an over zealous researcher if you point out the hiccups of others–this will only help you refine your research and prove that you have examined the mistakes of others so you avoid replicating them in your study!

ERROR THREE: Eenie Meenie Miney Criteria

Consider everything when it comes to your inclusion and exclusion criteria! This is where your experimental design will be derived and how you will sample. You may have to analyze the sample size and do a few pre-tests to be certain the sample size selected is appropriate and holds enough power for a good result. Be sure to discuss the exact criteria used to include certain subjects. This will help when you come back to analyze your results and help other researchers when they come to examine or replicate your study. Other researchers may identify a group or criteria that you excluded without you realizing that you did. And let’s face it, we are human. The more perspective the better. And the more details you invest into your criteria, the more others can ascertain not only your study’s validity, but if you really went far enough with your considerations. Push yourself to write even the most obvious criteria, like age group or gender, but move beyond that. Do these people have a stable home? Are they married? Any history of a particular disease? How about mental health? Job security? Type of job? What makes them excluded from the study? Really get into these details–those evaluating your research in the end want to have faith in the person who did the research. And in defining your inclusion and exclusion criteria, you will naturally come to assess any bias in your methodology, samples, measurements, or evaluations. So get into the nitty-gritty from the getgo!

ERROR FOUR: Your Lingo is Bingo

Remember that thing called the literature review. Usually the initial literature review derives key terminology that researchers list in their abstract. Those are important. Terminology is a structure, and from that structure is meaning. If words are being used and their entire implication is not understood by the researcher, the entire study can be a linguistic disaster. Be sure to evaluate and establish your key terminology, and define these according to any historical use of the word in relation to the subject at hand. If your term is a slight deviation from the past, note this and give a hearty justification as to why. Can a single study prove that a hypothesis is universally true? No. Unless it is a divine study that has omniscient powers, it can go so far as proving the null or supporting a hypothesis in that particular context. Calling out a universal truth can be dangerous if people start to apply your method or solution and come out with different results. Being cautious of your wording in regards to your hypothesis is kind of a big deal, and in a way, it is humbling. Just as the accuracy of numbers is important, the words you use to describe occurrences is equally important. Understanding the power of language will complement the scientific process altogether. Review some terminology related to the life of a researcher.

ERROR FIVE: Get Vetted before you get Taxed

This one is a big one that happens with novice researchers. Just as there is a pre-test, an assessment of a sample, and essentially checking out your design before you go to the big league of research, you need to get your heroes and critics to evaluate your research design. These mentors are your research Justice League action squad. These are typically professors, experts, or professionals in the field who can give you insight before you go all out with your research. In your research proposal, you are usually required to write about “the panel” who checked out your protocol and cleared it as ethical and sound. They may identify some gaps and spare you the time and effort of discovering these later on. So give them the chance to review, and most universities require them to approve your research anyways. If you want them to take your research seriously, take their time seriously. They are giving you the chance to present your ideas. When it comes to checking the relevance of your measurement tool, a statistician should be your best friend, or your panel of peers should refer you to someone with a background with such measurements. This is serious stuff–if your numbers come out saying something that is not true, not only can the entire study be compromised, but ethical dilemma’s can surface. Do not rely only on software. Machines are great but a touch of human is important to doing your due diligence.