STEP THREE

Create a shortlist of 4-5 potential journal articles

When it comes to creating a shortlist of 4-5 potential journal articles, you could do this in anywhere from 30 minutes to 2 hours, depending on the number of results that are returned from your search. However, since choosing the main journal article is so important in Route #1: Replication-based dissertations, we would advise spending 1-2 days doing this, just so that you have sufficient time to work through the available articles in your search. Even though you are not choosing a particular route at this time (i.e., Route A: Duplication, Route B: Generalisation or Route C: Extension), these routes still influence how you shortlist potential articles. All we would suggest is that you don't eliminate articles from your shortlist at this stage based on a particular route, which is something that we focus on later in the article, STAGE TWO: Choosing your route.

To create this shortlist of 4-5 potential journal articles, follow these five steps: (STEP A) Eliminate all publications on the list that are not journal articles; (STEP B) Eliminate those journal articles on the list that do not interest you from a personal perspective; (STEP C) Eliminate those journal articles on the list that are not appropriate for Route #1: Replication-based dissertations; (STEP D) Briefly evaluate the remaining articles, pulling out their main characteristics; and (STEP E) Choose a shortlist of 4-5 journal articles that interest you the most. Each of these steps is discussed in turn:

STEP A
Eliminate all publications on the list that are not journal articles

Since you're only looking for journal articles, eliminate all books, conference papers, and theses or dissertations that appear in your search results. For example, from the 75 results returned from our search, only 24 of these were journal articles, which we've listed in the diagram below.

STEP B
Eliminate those journal articles on the list that do not interest you from a personal perspective

Skim through the titles and abstracts of the journal articles, and simply eliminate those which you don't like the sound of (i.e., your gut feeling is important). For example, looking at the list of 24 possible journal articles in our list, we may choose to eliminate the following journal articles [NOTE: our reasons for rejecting the articles are fictitious, but should give you a sense of the types of reasons worth considering]:

Journal article #1 in the list
Quality assurance labels as drivers of customer loyalty in the case of traditional food products
In Good Quality and Preference

From the title alone, we decided that we had no interest in studying quality assurance labels or traditional food products. Therefore, trying to duplicate this research would have been boring, and trying to generalize or extend it would have probably required us to still examine at least one of these aspects of the original research (i.e., quality assurance labels as a driver of customer loyalty, or traditional food products as the setting/context for the dissertation). It took all of 5-10 seconds to rule this article out.

Journal article #3 in the list
Time-varying effects in the analysis of customer loyalty: A case study of insurance
In Expert Systems with Applications

From the title, it wasn't particularly obvious what was meant by time-varying effects in the analysis of customer loyalty. Whilst this is not a reason in itself to eliminate the article, a skim of the abstract revealed words that suggested the article had a heavy focus on statistical analysis (e.g., words like survival analysis, fixed and fluctuating model parameters, etc.). This felt like too much of a focus on data analysis, especially in statistical techniques that we were less familiar with [NOTE: as mentioned, this is a fictitious example. We do know our stats!].

You should be able to quickly reduce your list when eliminating journal articles that do not interest you from a personal or practical perspective.

STEP C
Eliminate those journal articles on the list that are not appropriate for Route #1: Replication-based dissertations

This is a quantitative dissertation. If you're duplicating, the main journal articles needs to be a quantitative piece of research. If you're generalizing or extending, the main journal article could be a piece of qualitative or mixed methods research, but this would then negate many of the advantages associated with taking on Route #1: Replication-based dissertations (i.e., these advantages include a strong platform to build on, such as well-developed hypotheses, a clearly set out research strategy, and comparable results). Therefore, we would suggest eliminating those journal articles on the list that have adopted a qualitative research design, and even those of a mixed methods research design if the quantitative component of such journal articles is not the dominant one.

STEP D
Briefly evaluate the remaining articles, pulling out their main characteristics

Steps 1-3 should have helped you to reduce your list down to a small number of journal articles. Now it is time to briefly evaluate the remaining journal articles to determine which ones to eliminate further, and which require more investigation. When working through the remaining journal articles, you need to consider (a) the number of sub-fields and constructs you'll have to consider, (b) the complexity of the data analysis you may need to understand and replicate, (c) the characteristics of the research strategy that you might have to follow or build on, and (d) the quality and specificity of the journal where the article is published:

• The number of sub-fields and constructs you'll have to consider

  Some research focuses on a single sub-field and a very small number of constructs (i.e., between two and three), whilst others is much broader. You can often determine the number of sub-fields and constructs from the title alone. Compare the following two journal articles on our list, the first with five such sub-fields/constructs, the second with just two:

  Journal article #7 in the list
  Interrelationships between physical environment quality [#1], personal interaction quality [#2], satisfaction [#3] and behavioural intentions [#4] in relation to customer loyalty [#5]: The case of Kinmen's bed and breakfast industry
  In Asia Pacific Journal of Tourism Research
  

  Journal article #22 in the list
  The effect of switching barriers types [#1] on customer loyalty [#2]
  In International Review of Business Research Papers

  The more sub-fields and constructs that are being addressed by a journal article, the greater the potential workload that you will have when it comes to performing your literature review. Whilst this should not be a reason in and of itself to reject a journal article, you do need to consider potential workload when comparing journal articles. Also, since you may have focused on a single sub-field in your search, ask yourself: Am I interested in the other sub-fields explored in this journal article? For example, in journal article #7 above, we would have to be interested in the sub-fields/constructs of [#1] physical environment quality, [#2] personal interaction quality, [#3] satisfaction and [#4] behavioural intentions, in addition to the main sub-field we were interested in relation to [#5] customer loyalty.
  

• The complexity of the data analysis you may need to understand and replicate

  Even though you are taking on a quantitative dissertation, this does not mean that you are a statistics expert. In fact, most undergraduate and master's level students that take on quantitative dissertations will have very little statistics knowledge. Now we do help you to select the appropriate statistical tests to use in your dissertation, statistically analyse your data using SPSS, and interpret the output from such analysis in the Data Analysis section of Lærd Dissertation. We even do this assuming little or no knowledge of statistics and SPSS, using straightforward, non-technical language. However, there is no question that some students find statistics harder than others. As such, it is important at this stage that you evaluate the complexity of the data analysis used in the journal articles remaining on your list. After all, not only will you need to understand the statistical analysis performed in the journal article, but in taking on a Route #1: Replication-based dissertation, you may have to replicate such analysis to some extent.

  The data analysis techniques used will often be detailed in the abstract, but if not, certainly in the methodology section of the journal article. As a general rule of thumb, if you are someone that finds statistics more challenging than most, avoid journal articles that mention statistical tests such as structural equation modelling (SEM) or partial least squares (PLS). You may even want to avoid statistical tests such as principal components analysis (PCA), factor analysis, logistic regression, and loglinear analysis, although we do have articles that will walk you through these statistical tests, step-by-step, in the Data Analysis section of Lærd Dissertation. As such, you may want to take a quick look through these statistical tests first to determine what is involve before eliminating journal articles with such statistical tests from your list. Just as a word of warning, the Data Analysis section and the writing up of such analysis in the Results chapter are the only two parts of Lærd Dissertation that are not free to access. However, we've made it as low cost as possible! Read on or Take the tour.

• The characteristics of the research strategy that you might have to follow or build on

  The research strategy adopted in a journal article refers to the approach taken by the authors towards (a) designing the research (i.e., whether a descriptive, experimental, quasi-experimental, or relationship-based research design was used), (b) sampling the population of interest (i.e., whether a probability or non-probability sampling strategy), (c) the research methods that were used to collect data (i.e., a survey, structured observation or interviews, data sets, laboratory methods, etc.), as well as (d) the data analysis techniques used, as we touch on in the previous bullet. You can learn more about these components of research strategy in the Fundamentals part of the site. However, the main point is not to evaluate such components of research strategy in any depth at this stage, but rather get a sense of what was involved in the journal articles remaining on your list, and assess whether you should eliminate any further journal articles based on such components. For example, if the journal article used structured interviews or observation, but you would prefer to deal with an existing data set rather than people, you may choose to eliminate the journal article on this basis. Alternately, the abstract of a journal article may suggest that too much work is involved in a replication-based dissertation. Take the following journal article in our list:

  Journal article #8 in the list
  Creating customer loyalty through service customization
  In European Journal of Marketing

  We may have eliminated this journal article from the list because the study involved two large-scale studies in different service industries. At the undergraduate and master's level, it would simply be too much work to (a) carry out two studies and (b) replicate studies that were particularly large. Nonetheless, unless there are particular aspects of research strategy that you do not feel comfortable with, we would not worry too much about the research strategy adopted in the journal article at this stage since this is something that you will inevitably look at in more detail when you read the entire journal articles on your shortlist, which we discuss later in STEP FOUR of the next section. But it is something to bear in mind.

• The quality and specificity of the journal where the article is published

  Not all journals are 'quality' journals. It's bad to have to say it, but some journal articles are of a poor quality. Journal quality is important because if an article has too many flaws from the start, you could simply be trying to replicate/build on a flawed study. In some cases, this is pointless because it would be better to address the problem from scratch. Also, you want to be able to build on a reasonably strong journal article, (a) where you wouldn't necessarily expect duplication to lead to different results, or where you are focusing on repairing specific limitations within the study, or (b) that provides a more stable platform to make generalisations or extensions. Your supervisor should be able to tell you which are the better journals in a particular field.