Writing in the Disciplines Knowledge Base

Writing Discussions

The discussion section is a framing section, like the Introduction, which returns to the problem statement set up in the Introduction. Reread your introduction carefully before writing the discussion; you will discuss how the research question has been answered by the new data and then show how the field’s knowledge has been changed by the addition of this information. While the introduction starts generally and narrows down to the specific hypothesis, the discussion starts with the interpretation of the results, then moves outwards to contextualize these findings in the general field.

The Discussion section is sort of an odd beast because it is here where you speculate, but must avoid rambling, guessing, or making logical leaps beyond what is reasonably supported for your data.  The solution that has evolved over time is to set up the Discussion section as a “dialogue” between Results  — your data in conversation with other people’s data. In other words, for every experimental result you want to talk about, you find results/models/conclusions from other publications bearing the relationship to your result that you want the reader to understand.

• Claim — add new information to what is already known — “we are the first to show”
  • no citations because this is the first time this data has been reported
• Corroborate — support what is already known — “similar to /same as X”
• Clarify — extend or refine what is already known — “because X, also Y” or “because X, not Y”
• Conflict — counter or contradict what is already  known — “contrary to”

This is how the new data you’ve generated is “situated” in the field — by your careful placement of what is new against that which is already known.  Results can take the form of data, hypotheses, models, definitions, formulas, etc.  (I imagine the Results section like a dance with swords — sometimes you are engaging your partner with the pointy end and sometimes you are gliding alongside them).

Steps to Crafting a Discussion Section

1. Briefly restate RQ and main result/s.
2. Situate current research findings through “dialogue” model — speculate on outcome.
3. Discuss relevant limitations.
4. Provide suggestions for future research or application.

Style Note: Overall, Discussion sections don’t differ according to method. In other words, Discussions in quantitative research and qualitative research are quite similar.

Example Step 1: Quantitative

As recently as the 1980’s, comic book heroes, high fantasy, and science fiction—media interests typically associated with geeks—were considered strange, unpopular, and in many cases taboo. In 2014, these same markers of geek culture are box office smashes, multi-billion dollar industries, and a wide-reaching counterculture with its own brands, fashion trends, and celebrities.

We sought to better understand the phenomenon of geek culture primarily at the individual level—that is, to understand why a given individual would choose to engage in geek culture. We developed and validated two scales to measure two major components of geek culture: engagement and identity. We also proposed and found mixed evidence for each of three models of geek cultural engagement. We review these findings below.

Analysis: The blue text restates the research questions while the orange provides a sentence summary of the major outcomes. The purple paragraph briefly reviews the topic and significance of the topic — while not a required step, it makes sense in this report because the study included 7 tests, each with its own subgoals. That’s a lot of information between the Introduction and Discussion. In addition, each test has its own Results and mini-Discussion section; thus, the final Discussion section of this report synthesizes the findings of all of the studies.

Example Step 1: Qualitative

This study investigated the socio-psychological factors influencing adult vaccination uptake in the UK. Our results suggest that the public have no general concept of adult immunization, as they have for childhood immunization. Instead, their beliefs and attitudes are vaccine-specific and in some cases age-specific. Participants classified influenza and tetanus, and to a lesser extent pneumonia, as severe diseases. Consistent with their disease appraisal, participants felt that the influenza, tetanus and pneumococcal vaccines were important. Understandably, few participants had heard of the hepatitis vaccines or knew about hepatitis, as both hepatitis A and B are uncommon in the UK. Participants generally associated influenza and pneumonia vaccines with older age, tetanus with adolescence and MMR with childhood.

The blue text restates the research questions while the orange provides a summary of the major outcomes.

As the Style Note above pointed out, overall, there isn’t much difference between quantitative and qualitative.

Examples of Relationships

Claim — “first to show”

The GCES is the first measure of its kind to focus specifically on the geek subculture. The GCES shows excellent reliability and construct validity. It adequately distinguishes self-identified populations (e.g., Dragon*Con attendees) and correlates positively with actual behavior (Study 3). It captures nuances of geek engagement that are not apparent to naïve observers (Study 2). Despite several of its factors having only two items, it presents a stable factor structure, with the majority of its subscales showing appropriate reliability. The possible exceptions are the Puppetry/Robotics and Roleplaying subscales—despite their face validity, these subscales contain only two items and show relatively low correlations. However, these subscale scores remain correlated to the other subscales and to geek engagement as a whole, and inclusion of their items in the full scale score does little to harm the overall reliability of this measure.

Corroborate — “similar to/supports”

The results of Study 5 appear to support the belongingness hypothesis. Geeks appear to form the strongest ties with those sharing similar specific geek interests, but also are more likely to form strong ties with other geeks who have similar interests. Although these results should be interpreted with caution due to noncompliance issues (i.e., the majority of participants not listing the requested number of alters), this general geek homophily is consistent with past research conceptualizing geek culture as using geek interests as social currency [8]. Future research should examine whether this homophily results from shared norms, beliefs, and values between the different fandoms in geek culture.

Clarify — “this, and also that” or “this, but maybe not that”

Our predictions were only partially supported with regards to the openness, creativity, and need for stimulation correlated with geek engagement. Of the variables tested, geek engagement appears to be predicted primarily through creativity and its correlates. Neither need for cognition nor sensation seeking appear to play a role. In addition, the negative relationship with crystallized intelligence conflicts with the common belief that geeks are more intelligent than non-geeks. This may partially be a result of using a brief self-report scale of intelligence. Although the Shipley Institute of Living Scale [75] is a well-established brief measure of intelligence, a more in-depth IQ battery such as the WAIS-IV [78] may be needed to detect more nuanced relationships between geek engagement and intelligence. Although surprising, geek engagement’s negative relationship with intelligence coupled with its positive relationships to dissociative and schizotypal symptoms are consistent with DeYoung’s [79] conception of openness to experience as a paradoxical simplex in which intelligence and apophenia (a trait similar to schizotypy) are both related to openness but also different from each other. These results suggest that those who are high in geek engagement are on the apophenia (rather than intelligence) region of the simplex. This could explain the common belief that geeks are more intelligent because their openness resembles that of individuals high in the intellect region of the circumplex. However, their tendency toward apothenia may determine why some individuals high in openness gravitate toward geek activities while others do not.

Conflict — “because of this, not that”

Studies 1, 4, and 5 addressed the belongingness hypothesis, which predicts that individuals will engage in geek culture to fulfill belongingness needs. Belongingness has long been considered a basic need [23] and Self-Determination Theory [24] posits that much of human motivation is driven by basic needs related to belongingness (e.g., relatedness). Study 1 provided little support for this hypothesis, as the relationships subscale of the BPN scale showed no relationship to geek engagement. This implies that those high in geek engagement are neither more nor less likely to have fulfilled their belongingness needs than those who are low in geek engagement.

Point out relevant limitations

All studies have limitations, aspects of the research that could not be controlled for. Some, such as self-report, are so well known in psychology that they are mentioned briefly or only with respect to a specific result. Sometimes, broader limitations impact the whole study and the responsible researcher must point these out or risk losing credibility. Broader impacts still tend to be in their own labeled section of the paper in psychology (and the social and behavioral sciences overall).

Given this, hyper-focus on limitations also damages credibility. You simply don’t need paragraph long confessions of every variable outside your control. No one has unlimited time or infinite resources. Instead, pay tribute to limitations that most likely impacted methodology or results. For example, students often have more limited demographic diversity in their participant pool because they have access to fewer people or material resources. Acknowledge this, and when appropriate, consider shifting the focus of a report to the demographic. For example, if your survey reached 37 people, 32 of whom identified as female, then shift the research question to account for this by making the research about females or split the analysis into the 32 females and 5 males.

Finally, limitations can lead to suggestions for future research! Research prose in the last 20 years or so has evolved to be more cohesive, to have a bit more “flow”. Thus, limitations are frequently pitched as opportunities for more work.

Examples

The Geek Culture article includes limitations in 2 ways: in the Discussion of specific findings and in a section following the Discussion called “Limitations and Future Directions”. The Vaccine in Adults study provides limitations in a sub-section of the Discussion called “Limitations”.

The Geek Culture Engagement Scale (GCES) and the Geek Identity Scale (GIS)

The GCES is the first measure of its kind to focus specifically on the geek subculture. The GCES shows excellent reliability and construct validity. It adequately distinguishes self-identified populations (e.g., Dragon*Con attendees) and correlates positively with actual behavior (Study 3). It captures nuances of geek engagement that are not apparent to naïve observers (Study 2)....

There are important limitations to the interpretations that can be drawn from the GCES. First, because we used major geek conventions to generate the list of activities for the scale, this scale may fail to capture more marginalized geek activities that are not represented at a large convention. Second, because we conceptualized geek culture engagement as involvement in multiple geek activities, this scale may not capture geek “specialists,” or persons engaging intensely in only one geek activity (e.g., an avid Trekkie who only devotes his time to Star Trek). Although specialists were relatively rare in our samples, they did appear to differ from other geeks in terms of Big Five personality variables—especially agreeableness, where specialists reported relatively low levels. Thus, the GCES speaks best to generalist geeks, and caution should be used when specifically studying specialist geeks. However, homophily between geeks with specific interests (Study 5) only emerged when controlling for intercorrelation between subscales, and the Geek Identity Scale (GIS) correlates positively with the full scale GCES, implying that identification as a geek intensifies as one is engaged in more and more geek activities. Although geek specialists may exist, these persons may identify less with geek culture per se, and identify more strongly with their chosen fandom, as evidenced by their lower overall score on geek identity.

Limitations and Future Directions

In this paper we have only begun to explore the reasons people engage in geek culture. As we state up front, this is a beginning rather than the last word on the topic. We have relied heavily (although not exclusively) on correlational, self-report data to examine the plausibility of the theories posed above. Experimental, developmental or experience sampling methods would be ideal to more definitively test each of the hypotheses proposed in this paper. We have foregone more complex mediational analyses that will eventually be required to provide a definitive test of the mechanisms we have proposed here. We also have not conducted research using other enthusiasts as a comparison group; research comparing geeks to other groups containing like-minded individuals (e.g., football fans) will be needed to determine whether these relationships are exclusive to geeks. Finally, we have focused on these hypotheses at an individual level. Cultural level work exploring major cultural events and demographic information is needed to examine these hypotheses, as geek engagement is a cultural trend as well as an individual behavior.

In addition, there is a strong reliance on MTurk as the source for most of the samples used (with the exception of our sample from Dragon*Con). Although there is little reason to expect MTurkers to differ appreciably from the wider population [42,43], use of a wider range of samples in future work would be useful.

Finally, even within geek engagement, more work needs to be done to discern what makes these media interests part of geek culture. For example, what role does escapism play in geek culture? Is escapism the common factor that attracts geeks to a new franchise? Does the appeal lie in some element of “magic” or controlling the uncontrollable? Do the media need to include some sort of “special” individual who has extraordinary powers or has been chosen for some quest? Additionally, geeks are theorized to share social norms, values, and customs in addition to common interests [6–8]. Work using techniques from cultural psychology or sociology may help to illuminate these elements of geek culture.

Analysis: Limitations are indicated in purple and suggestions for future research in green. In the second excerpt, bolded purple / green pairs indicate limitation statements which lead directly to suggestions for improved research.

Suggest future research or subsequent action

Thus far, the Discussion section has interpreted study outcomes with respect to the field. After all of this work, the reader wants to know “What should happen next?” Provide the answer in one of two forms: suggestions for future research or suggestions for specific actions.

All research reports should include suggestions for subsequent research. Future research has to be connected to the study, so emerges from the Discussion relationships, especially Claim, Clarify, and Conflict. Not every outcome merits future research, but when it does, the current pattern is to include the suggestion immediately after the outcome has been discussed.

A striking and novel psychological finding of this study is that previous experiences related to injections, particularly during childhood, had both a positive and a negative influence on vaccination uptake. One-third of influenza non-vaccinators reported having had a traumatic experience with vaccines, injections or medication in the past, which they stated had influenced their decision to not vaccinate against influenza. This resonates with previous research which showed that painful neonatal experiences such as circumcision or Heel Stick Capillary Blood sampling in neonates can magnify the experience of pain later in life [43], [44]. Similarly, some tetanus vaccinators recalled that the memory of their mother's warnings about the danger of contracting tetanus in childhood had influenced their decision to have a tetanus booster. These findings suggest that pain (caused by needles) and fear (of contracting a severe disease) during childhood could become both a potent vaccination deterrent or enabler. More research in this area is needed.

• Page 2 — Introduction
• Page 3 — Methods
• Page 4 — Results
• Page 5 — Discussion
• Page 6 — Action Plan and Outlines