What is Theorizing

Research Project 1 (RP1) — U8

Andy Weeger

Neu-Ulm University of Applied Sciences

September 3, 2026

Opening Remarks

Building Explanations

Theory is a noun.
Theorizing is a verb.

Last session was about recognizing a finished theory when you see one in the literature. This session is about the messier process that produces one: how you get from an observation, something one of your practitioners said, something a pattern in your review keeps surfacing, to a general, testable explanation of it.

Learning Outcomes

By the end of this unit, you will be able to:

• Move a circular or overly specific explanation of an observation to a general, falsifiable one by naming its underlying mechanism.
• Propose a competing explanation for the same observation, and name a context in which the two would predict different outcomes.
• Judge a theoretical account against seven quality checks: insight, testability, evidence, parsimony, logic, boundary conditions, and implications.

Theorizing

Definition

Theorizing is the application or development of theoretical arguments to make sense of a real account, for example an observed phenomenon (Recker, 2021).

• Theorizing can be inductive, deductive, or abductive.
• Theorizing can be dependent on data analysis, creative thinking, inspiration, or good luck.

Principles

Theory is always a simplification.

It should not be more complex than the phenomenon it investigates.

A strong theory is an idealization.

For instance, the idea of rational human decision-making.

General Approaches

From theory to data
Start with a theory, modify or extend it, test the predictions

From data to theory
Start with data, check if there is a theory that can explain what you observe, develop a new theoretical account

From Theory to Data

Typically called the “traditional” scientific process.

• You start with a theory.
• You develop a modification or an extension of it.
• Then you collect data specifically to test or falsify the predictions.

From Data to Theory

Often equated with inductive and interpretive research.

• You start with data, with a study of something that is “really happening.”
• You examine whether there is a theory to explain what you observe.
• You start developing a novel theoretical account, “grounded” in the data.

Theorizing Approaches

Building theory,
testing theory,
and extending theory.

The approaches are not mutually exclusive, and researchers often combine elements from multiple approaches depending on the research context.

The choice depends on the research question, available data, and the goals of the study.

Theory Building

Building theory involves the development of new theoretical frameworks or explanatory models to understand a phenomenon. This process often starts with empirical observations or data collection. Researchers use inductive or abductive approaches to identify patterns, relationships, and underlying mechanisms, and formulate conceptual frameworks that capture these patterns. Building theory is particularly useful when there is a lack of established theoretical foundations for a specific area of study (Mueller & Urbach, 2017).

Theory Testing

Testing theory involves subjecting existing theoretical frameworks or models to empirical scrutiny to assess their validity and predictive power. Researchers use deductive testing to generate hypotheses based on established theories and then gather data to confirm or refute these hypotheses. Inductive testing uses empirical observations to challenge or refine existing theories.

Theory Extending

Extending theory refers to enhancing or expanding existing theoretical frameworks to refine and broaden their applicability, accommodating new empirical findings and addressing limitations. Researchers can extend theory by

• introducing new concepts, dimensions, or variables to an existing theory (conceptual extension);
• adding new relationships, mechanisms, or components to an existing framework (theoretical extension); or
• integrating insights and concepts from other disciplines (interdisciplinary extension).

Quality Checks

Recker (2021), Mueller & Urbach (2017), and others propose that you have a good theory when you can answer the following questions:

• Is your account insightful, challenging, perhaps surprising, and, importantly, does it seem to make sense?
• Is your account (your arguments) testable (falsifiable)?
• Do you have convincing evidence to support your account?
• Is your account parsimonious?
• Are the arguments logical?
• What can you say about the boundary conditions of the theory?
• What are the implications of your theory?

Hold on to the boundary-conditions question in particular. When you reach the synthesis stage of your review and meet studies that appear to contradict one another, apply Bacharach (1989)’s criterion by reconciling them through the contingent factors that explain why one study found X and another found not-X, rather than averaging across them or dismissing the disagreement as noise. That is exactly the move you will need to make in SLR IV: Analysis and Synthesis: treat a contradiction in the literature as a clue to a boundary condition someone has not yet named, rather than a problem to explain away.

Worked Example

Start with an observation.

• Think about being in college. You’re in class, and the guy next to you, who is obviously a football player, says an unbelievably dumb thing in class. Why?¹
• Initial theory: Football players are dumb.

Theories should be about classes of things, that is, more general.²

• New theory: Athletes are dumb.

Theories should be explanatory, and free of circular arguments, since circularity prevents theories from being falsifiable.

• Dumbness cannot be directly observed or measured. The only way we can know if people are dumb is by what they say and do. So we say that they say dumb things because they are dumb, a circular argument.
• New theory: To be a good athlete requires lots of practice time; being smart in class also requires study time. Amount of time is limited, so practicing a sport means less studying, which means being less smart in class.

A good theory is general enough to generate implications for other groups and contexts, all of which serve as potential tests of the theory, that is, the theory is fertile.

• We can shift the focus from an enduring property of a class of people (athletes) to a mechanism, which lets us apply the same reasoning to other people and situations.
• New theory: There is limited time in a day, so when a person engages in a very time-consuming activity, such as athletics, it takes away from other very time-consuming activities, such as studying (Limited Time Theory).

Often, there are also other explanations:

• Everyone has a need to excel in one area. Achieving excellence in any one area is enough to satisfy this need. Football players satisfy their need for accomplishment through football, so they are not motivated to be smart in class (Excellence Theory).
• We are jealous of others’ success. When we are jealous, we subconsciously lower our evaluation of that person’s performance in other areas. So we think football players ask dumb questions (Jealousy Theory).

We need to study different contexts to see how our theories explain observations.

• Depending on the context, each theory leads to other expectations, e.g.,
  • Do football players ask dumb questions out of season?
  • Will athletes who do not look like athletes ask dumb questions?
• The expectations of our theory lead to hypotheses that can be tested by empirical research.

Worked Example: Streaming Churn

• Observation: subscriber churn spikes every Sunday night.
• Naive: “Sunday users just don’t stick around.”
• Mechanism: finishing a binge-watched series ends that week’s reason to keep the subscription open.
• Competing: weekend viewing degrades recommendation quality, and weaker recommendations drive churn.
• Discriminating test: does churn track series completion, or measured recommendation accuracy?

A streaming company notices subscriber churn spikes every Sunday night.

The naive explanation: “Sunday users just don’t stick around,” circular, since it just restates the observation as a trait of the day.

Generalize and name the mechanism: finishing a binge-watched series ends that week’s reason to keep the subscription open. Completion drives the drop, independent of which day it happens to fall on (Completion Theory).

A competing explanation: the recommendation engine trains mostly on weekday viewing behavior, so weekend-heavy usage gets weaker recommendations, and weak recommendations drive churn (Recommendation Theory).

The discriminating test: Completion Theory predicts churn among anyone who just finished a series, on any day, regardless of recommendation quality. Recommendation Theory predicts churn should track measured recommendation accuracy, regardless of whether a series was just completed. Checking churn among binge-completers who still received accurate recommendations separates the two.

The sequence is identical to the one you just ran on a college classroom observation: name the mechanism, generalize it, then locate where two accounts diverge.

Team Exercise

Team Theorizing

Pick one observation from the industry inputs. Theorize it through the same sequence.

1. State the initial, likely circular or too-specific, explanation, as a practitioner might state it.
2. Generalize it to a class of things.
3. Remove the circularity, name the mechanism.
4. Propose at least one competing explanation for the same observation.
5. Name a context in which the two explanations would predict different outcomes, your discriminating test.

15:00

Debrief in plenum: for each team, ask specifically for step 5, the discriminating context. A pair of competing explanations that make identical predictions in every context you can think of redescribes the same idea twice rather than doing useful theoretical work. This is the same standard Lave & March (1993)’s Limited Time, Excellence, and Jealousy theories were held to: each one is falsifiable against a different set of contexts.

The Theorizing Loop

• Observation: something a practitioner said, or a pattern your review keeps surfacing.
• Generalize: move from a specific instance to a class of things.
• Mechanism: replace a circular label with the process that actually produces the outcome.
• Competing explanation: propose at least one alternative account of the same observation.
• Discriminating test: name the context where the two accounts predict different outcomes.

You ran this loop twice today, once on a college classroom observation, once on a streaming service’s churn data, and your team then ran it a third time on your own industry input. The loop does not stop here: any claim in your Map of Perspectives or Problem Framing Brief that reads as “X causes Y” deserves the same five steps, generalized, given a named mechanism, checked against a competing account, and tested against a context where the two would diverge.

Homework

Read Hund et al. (2021) and make notes on the following questions:

• How do the authors theorize? From theory to data, or from data to theory?
• What is their theoretical contribution to the existing body of knowledge?
• What are the building blocks of their theory (constructs, propositions, logic, boundary conditions)?

Q&A

Footnotes

1. The example draws from Lave & March (1993).

2. Theories that are too narrow and specific (low-level theories) are not very interesting, even if they are correct.