Good Research Questions

Research Project 1 (RP1) — U6

Andy Weeger

Neu-Ulm University of Applied Sciences

September 3, 2026

Opening Remarks

Question Viability

Once you have identified an effective topic, you need to formulate research questions and propose a plan to address them.

This challenge is much more difficult than learning methods and theories, largely because it is undefined and highly contextual, without a fixed structure to rely on (Recker, 2021). The research question is the central element that shapes, and gets refined throughout, your whole review.

Learning Outcomes

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

• Diagnose a research question against four failure symptoms and five quality criteria, and revise a question that fails either.
• Motivate a research question through the four-step script, problem domain, phenomenon, knowledge gap, research question, synthesizing insight from practice and from literature into one logical chain.
• Choose between gap-spotting and problematization as strategies for generating a question, and justify which strategy your own tension calls for.

Main Problem

Four Symptoms

Recker (2021) outlines four primary problems that are indicative of a research question being inadequate.

The “monologuing” problem, the “so what” problem, the “solving the world” problem, and the “multitude” problem.

• The “monologuing” problem: you cannot state what research question you are tackling unless you engage in a five-minute monologue, you have not grasped the essence of the problem yet.
• The “so what” problem: you cannot tell why answering the research question matters to anyone.
• The “solving the world” problem: your question has value, but cannot be answered given your resource constraints.
• The “multitude” problem: you ask (too) many questions instead of one.

Further categories of inappropriate research questions are (Recker, 2021):

• Obvious questions: “Are there challenges in using information technology?” Of course there are. Obvious questions have answers to which everyone would agree.
• Irrelevant questions: “What is the influence of weather on the salaries of technology professionals?” There is no reason to believe there is any influence whatsoever.
• Absurd questions: “Is the earth flat after all?” Absurd questions have answers to which everyone would agree.
• Definitional questions: “Is technology conflict characterised by disagreement?” The answer is simply a matter of creating a concept that says it does. A definition is a form of description. It does not count as research.
• Affirmation questions: “Can a decision-support tool be developed to facilitate decision-making for senior retail executives?” Yes. Trivially.

Good Research Questions

Criteria

A good research question should be:

• Clear and focused: the question should clearly state what you need to do.
• Not too broad and not too narrow: it must be possible to answer the question within the constraints of your review.
• Not too easy to answer: the answers should not be obvious or affirmative.
• Researchable: you must have access to the literature required to answer the question.
• Analytical rather than descriptive: the question should allow you to produce an analysis of the problem rather than a simple description of it.

Once you have identified an important phenomenon, you need to engage with literature to identify problems with existing knowledge. This helps you narrow down your topic and create a good research question.

Research questions are typically one of two types, based on the issues they address (Recker, 2021):

1. “What,” “who,” and “where” questions tend to focus on issues we seek to explore or describe because little knowledge exists about them.
2. “How” and “why” questions are explanatory, seeking to answer questions about the causal mechanisms at work in a particular phenomenon.

Guiding Questions

Recker (2021, pp. 35–36) proposes a number of guiding questions that can help you find a good research question, e.g.:

• Do you know in which field of research your research question resides?
• Do you have a firm understanding of the body of knowledge in the field?
• What are important open research questions or unsolved problems in the field that scientists agree on?
• What areas need further exploration?
• Could your study fill an important gap in knowledge?
• Has your proposed study been done before? If so, is there room for improvement or expansion?
• Is the timing right for the question to be answered?
• Who would care about obtaining an answer to the question?

Discussion

Pick three of the guiding questions introduced before and answer them for your own emerging research question.

10:00

Exercise

Check the following research questions, do they meet the criteria of a good topic and RQ?

• What are consumer- and product-related antecedents to differences in the valuation of human- vs. algorithm-created digital products? (Rix et al., 2025)
• Whether and how do conversational style and reciprocation triggers influence user information disclosure? (Adam & Benlian, 2024)
• How do daily congruence and incongruence in the team-enacted versus individually needed use of agile practices affect developer well-being? (Benlian et al., 2025)
• How does the introduction of a substitutive decision-making AI system affect employees’ professional role identity? And how do employees adapt their professional role identity in response to these AI systems? (Strich et al., 2021)
• How can IT support be provided for reducing the problem-solving complexity of monitoring the structural and dynamic complexity of IT architectures in the context of a digital business strategy? (Widjaja & Gregory, 2020)

15:00

What are consumer- and product-related antecedents to differences in the valuation of human- vs. algorithm-created digital products? (Rix et al., 2025)

• Significance: very high, taps into AI/consumer behavior, timely for digital markets.
• Novelty: yes, “algorithm discount” is an emerging phenomenon.
• Curiosity: clear puzzle, why do consumers devalue algorithmic output?
• Scope: ambitious but focused, restricted to antecedents, not all possible outcomes.
• Actionability: strong, firms can adjust design and marketing strategies.
• Clarity: excellent, concise, no ambiguity.
• Answerability: yes, lends itself to experiments or surveys.
• Analytical focus: clear independent (consumer/product antecedents) and dependent (valuation differences) variables.

Whether and how do conversational style and reciprocation triggers influence user information disclosure? (Adam & Benlian, 2024)

• Significance: high, relates to privacy and chatbot/UX design.
• Novelty: yes, adds nuance (stylistic and social triggers) beyond generic “chatbot vs. form.”
• Curiosity: medium to high, “whether and how” signals uncertainty and depth.
• Scope: appropriately narrow, two levers affecting one key outcome.
• Actionability: clear, design implications for conversational agents.
• Clarity: minor flaw, “whether and how” could be split into two clearer sub-RQs.
• Answerability: yes, testable in lab or field experiments.
• Analytical focus: strong, well-specified factors and outcome.

How do daily congruence and incongruence in the team-enacted versus individually needed use of agile practices affect developer well-being? (Benlian et al., 2025)

• Significance: high, agile use and misalignment is a core IS management issue.
• Novelty: yes, fine-grained “daily congruence” adds a new perspective.
• Curiosity: high, counterintuitive potential, incongruence may harm or help.
• Scope: nicely bounded, one outcome (well-being), one context (agile).
• Actionability: very strong, managers can adapt agile routines.
• Clarity: long sentence but precise, terms are operationalizable.
• Answerability: yes, daily diary studies or panel data possible.
• Analytical focus: excellent, congruence/incongruence as predictor, well-being as outcome.

How does the introduction of a substitutive decision-making AI system affect employees’ professional role identity? And how do employees adapt their professional role identity in response to these AI systems? (Strich et al., 2021)

• Significance: very high, addresses identity threats in the future of work.
• Novelty: strong, few studies directly link AI substitution with identity adaptation.
• Curiosity: very strong, open question with human and organizational stakes.
• Scope: appropriately bounded to identity, not all socio-technical effects.
• Actionability: medium, insights more conceptual and theoretical than prescriptive.
• Clarity: excellent, two tightly worded sub-RQs.
• Answerability: qualitative or longitudinal research feasible.
• Analytical focus: good, causal mechanisms implied (AI introduction leads to identity shifts and adaptation).

How can IT support be provided for reducing the problem-solving complexity of monitoring the structural and dynamic complexity of IT architectures in the context of a digital business strategy? (Widjaja & Gregory, 2020)

• Significance: high, enterprise architecture complexity is a major IS challenge.
• Novelty: moderate to high, adds a design science angle to monitoring complexity.
• Curiosity: solid, “reducing problem-solving complexity” is compelling.
• Scope: quite broad, multiple forms of complexity, strategic context.
• Actionability: strong, artifact-oriented, prescriptive.
• Clarity: weaker, long and abstract phrasing, could be split into clearer parts.
• Answerability: yes, via artifact design and evaluation.
• Analytical focus: weaker than others, less explicit on variables and mechanisms.

Exercise

Exchange your current draft research question with another team.

Apply the five criteria and the four symptoms above to their draft, not yours. Note one thing that works and one thing you would revise.

15:00

Motivating Your Question

The Four-Step Script

Recker (2021, p. 36 ff) proposes a systematic four-step approach for motivating research questions that ensures a logical flow from broad context to specific inquiry.

Problem domain statement + specific phenomenon + knowledge gap or problem + research question

First, establish a problem domain statement that highlights an important business or organizational context, emphasizing the scale and significance of the domain while including quantifiable evidence when possible, dollar amounts, percentages of organizations affected, or market size metrics.

Second, identify a specific phenomenon by narrowing down from the broad domain to a particular issue or challenge, focusing on what makes this phenomenon especially important or problematic within the larger context.

Third, articulate the knowledge gap or problem by clearly stating what is wrong with current knowledge, a gap where we don’t know enough, inconsistencies where studies show conflicting results, outdated assumptions resting on unrealistic premises, or limited scope where existing research only covers part of the phenomenon.

Finally, formulate the research question as a logical conclusion flowing from the previous three steps, directly addressing the identified knowledge problem.

Example: Agile Practices

Problem domain statement

Organizations increasingly adopt agile software development practices to improve project outcomes and developer productivity, with 71% of organizations now using agile practices in their software development lifecycle.

Specific phenomenon

Despite this widespread adoption, agile projects continue to suffer from alarmingly high failure rates (50-96%), often attributed to human-related challenges where developers experience imbalances between the team-enacted use of agile practices and their individual needs.

Knowledge gap/problem

The literature to date has implicitly assumed that developer needs are perfectly reflected in team-enacted agile practices, overlooking the potential for incongruence between team practices and personal requirements.

Research questions

How do daily congruence and incongruence in the team-enacted versus individually needed use of agile practices affect developer well-being? Additionally, how pronounced is the influence of this congruence and incongruence among frequent (versus infrequent) team feedback-seeking developers? (Benlian et al., 2025)

This moves from the broad agile-adoption domain to the specific congruence phenomenon, identifies a clear knowledge gap, an untested assumption about alignment, and poses a research question that tests it directly.

Example: Social Engineering

Problem domain statement

Cybersecurity threats represent a critical organizational challenge, with digital system dependence creating vulnerabilities that extend beyond technical safeguards to human-centered attacks. Attackers increasingly bypass technical defenses by targeting the “easier to breach” human element through social engineering (SE) attacks. The urgency has intensified with AI advances enabling highly scalable, personalized, and convincing attacks such as automated phishing, deepfake-based impersonation, and adaptive social manipulation.

Specific phenomenon

Social engineering attacks represent a distinct socio-technical phenomenon where attackers exploit human vulnerabilities through social interaction rather than technical system breaches, relying on trust, deception, and psychological influence.

Knowledge gap/problem

Current understanding of social engineering suffers from both definitional fragmentation (50 unique definitions identified) and explanatory inconsistency, with dual-process theories providing conflicting evidence about whether peripheral or central processing creates vulnerability.

Research questions

• RQ1: What defines social engineering? Addresses the definitional fragmentation limiting theoretical development.
• RQ2: How does social engineering unfold? Resolves the explanatory contradictions by systematically synthesizing SE mechanisms across psychological, cognitive, affective, and contextual conditions.

This moves from the broad cybersecurity domain to the specific SE phenomenon, identifies clear knowledge problems (fragmentation and contradiction), and poses research questions that directly address these gaps.

Problematization

Weak Reasoning

“There is a gap in the literature” is the weakest possible reason to ask a question.

Ask which assumption might not hold instead. Alvesson & Sandberg (2011) name five types worth challenging.

• In-house: a specific claim within a school of thought, on its own terms.
• Root-metaphor: the field’s guiding image for a phenomenon.
• Paradigmatic: a tradition’s underlying ontological or epistemological commitments.
• Ideological: the political, moral, or interest-laden framing.
• Field: assumptions shared across competing schools of thought.

Everything so far has trained you to spot a problem with existing knowledge, a gap, and to build a question that fills it. Alvesson & Sandberg (2011) argue this default mode, gap-spotting, systematically under-produces interesting research, because it accepts the assumptions of the existing literature and merely extends its boundaries. You end up with incrementally more of the same conversation rather than a new one.

Problematization is the alternative: instead of asking “what hasn’t been studied yet,” you ask “what assumption is this literature making that might not hold?” (Alvesson & Sandberg, 2011) identify several types of assumptions worth challenging:

• In-house assumptions: challenging a specific claim or finding within a school of thought, on its own terms.
• Root-metaphor assumptions: challenging the underlying image or metaphor a field uses to think about a phenomenon (e.g., is “adoption” really the right metaphor for how organizations take up new technology, or does it presuppose a passivity that isn’t there?).
• Paradigmatic assumptions: challenging the deeper ontological or epistemological commitments a whole research tradition shares.
• Ideological assumptions: challenging the political, moral, or interest-laden assumptions embedded in how a phenomenon is framed.
• Field assumptions: challenging assumptions shared across multiple, otherwise competing, schools of thought.

Notice how this connects to what you already did in Synthesizing Insights: a genuine contradiction between two practitioner accounts is often evidence that a field’s assumption doesn’t hold in practice, which is exactly the kind of observation problematization asks you to build a question from, rather than a gap you noticed by counting what has and hasn’t been published.

Gap-Spotting vs. Problematization

	Gap-spotting	Problematization
Starting move	What hasn’t been studied yet?	What is this literature assuming that might not hold?
Relationship to literature	Extends its boundaries	Challenges its foundations
Typical yield	Incremental contribution	Potentially more novel, higher-risk contribution
Risk	Safe, but may be “so what”	Harder to execute, easier to overreach

Table 1: Two strategies for generating a research question, based on Alvesson & Sandberg (2011)

Neither strategy is automatically better. A gap-spotting question, done well, is still a legitimate and often more tractable contribution for a semester-length review. But a topic that only survives as a gap (“nobody has looked at X in context Y yet”) is fragile, someone else can fill the same gap. A topic that survives as a problematization (“the field assumes X, but our evidence suggests otherwise”) is harder to write, but harder to make redundant too.

Exercise

Take one of your research question candidates that tends to be gap-spotting and try to reframe it as a problematization.

Name the assumption the original literature is making, and state the reframed question in one sentence.

10:00

Homework

Prepare your own Problem Framing Brief and review the one of another team.

1. Write out the four-step motivation in full: problem domain, phenomenon, knowledge gap, research question (template see assignments).
2. Check your question against the five criteria and the four problem areas.
3. Upload your team’s current draft Problem Framing Brief to Moodle by Mon, 09.11, so the team reviewing you at the next session has time to read it beforehand.
4. Read the assigned peer-review readings (see Moodle) before the next session, to prepare for reviewing someone else’s framing.

Q&A