Latest / 5 Minute UX / Choosing Research Techniques: A Practical Guide
Transcript
- 0:07This lesson is Choosing Research Techniques: A Practical Guide.
- 0:11Let's get started.
- 0:18Before selecting techniques, you need to identify the four
- 0:21prerequisites required before selecting research techniques.
- 0:24Start by establishing a User Group Definition framework. This
- 0:27creates a defined set of user groups to prioritize for the
- 0:29project, which allows for focused recruiting efforts.
- 0:32Without this structure, your participant screening lacks
- 0:34direction. You must also conduct a Risk Assessment to identify
- 0:37high-risk areas lacking user insights. This helps the team
- 0:40understand where acting without data creates the most danger.
- 0:43Next, define Project Constraints including specific budget and
- 0:46timeline limits. These boundaries dictate what is
- 0:48actually feasible for your team. If direct user contact is
- 0:51impossible, create Existing Models verified with external
- 0:53sources like customer service. This mitigates the risk of
- 0:56incorrect assumptions when you cannot reach users directly.
- 0:59Once these prerequisites are in place, you can move to the four-
- 1:02step selection process.
- 1:08Once those prerequisites are set, you move into the four-step
- 1:11selection process that transforms constraints into a
- 1:13concrete plan. The first step is to define primary user groups,
- 1:15which means creating a framework that describes the main types of
- 1:18users you need to reach. This output gives you a defined set
- 1:21of groups to prioritize for the project, ensuring your
- 1:23recruiting efforts are focused rather than scattered. By
- 1:25narrowing your target audience early, you avoid the common
- 1:28mistake of trying to recruit everyone, which dilutes your
- 1:30insights and wastes valuable time. Next, you determine
- 1:32research goals and scope by deciding exactly what you want
- 1:35to learn from the study. You identify the primary topics to
- 1:37be covered, which results in clear research objectives and a
- 1:40topic coverage list. This step forces you to articulate
- 1:42specific questions, because every technique you choose must
- 1:45serve one of those defined goals. When your objectives are
- 1:47sharp, it becomes much easier to match them with the right
- 1:49methodological tools later in the process. The third step
- 1:52involves selecting techniques based on project phase and type,
- 1:54which requires applying specific decision logic to your
- 1:57situation. For task-based applications, you conduct user
- 1:59interviews before designing to understand needs, then perform
- 2:02usability testing on a prototype later. In contrast, for content
- 2:04sources, you start with contextual inquiry to observe
- 2:07behavior, then include a card sorting exercise later to
- 2:09validate structure. If you have room for at least two rounds of
- 2:11research, use the first round during the define or early
- 2:14design phase, and the second round before development starts.
- 2:16Finally, you plan logistics and resources to ensure the study
- 2:19can actually happen within your constraints. This means
- 2:21determining how to get participants through rigorous
- 2:23recruiting and screening, as well as setting appropriate
- 2:26participant compensation. You also need to determine space,
- 2:28equipment, and software needs based on whether the research is
- 2:31in-person or remote, along with deciding how information will be
- 2:34captured. This logistical plan covers recruitment,
- 2:36compensation, environment, and data capture tools, turning your
- 2:38strategic choices into an executable reality. With this
- 2:41framework in place, you can see how these steps apply to a
- 2:43specific task-based application scenario.
- 2:52Let’s apply this framework to a specific scenario where your
- 2:54team is designing a new task- based application with a two-
- 2:57round research window. Because the project type dictates the
- 2:59method, you’ll conduct user interviews in the first round to
- 3:02understand user needs before designing any interfaces. This
- 3:04happens during the Define phase or early in the Design phase,
- 3:06which means you’re gathering foundational insights while the
- 3:09scope is still flexible. For the second round, you’ll perform
- 3:11usability testing on a prototype before development starts,
- 3:14ensuring the design actually works for real users. This
- 3:16timing strategy validates your design decisions against actual
- 3:19behavior, so you don’t waste engineering resources on
- 3:21untested concepts. By applying the two-round timing strategy,
- 3:23you confirm usability before coding begins, creating a
- 3:26feedback loop that reduces risk. This sequence ensures insights
- 3:28inform design and validation confirms usability,
- 3:30demonstrating how technique selection aligns with project
- 3:33phase. Now that we’ve seen the ideal workflow, let’s look at
- 3:35what happens when those constraints shift or assumptions
- 3:38go unverified.
- 3:44Which of these three pitfalls is most likely to impact your
- 3:46current project's timeline? You might be using an unvalidated
- 3:49model, which introduces higher risk that your assumptions will
- 3:52be incorrect. The recovery is simple: verify the model with
- 3:54sources outside the project team, such as customer service
- 3:57or training groups. A verified model is better than having no
- 3:59model at all, so make that check before you proceed. Another
- 4:02common issue is a lack of familiarity with techniques,
- 4:04which can cause practitioners to struggle with determining the
- 4:07right research approach. To recover, refer to standard
- 4:09tables of commonly used research techniques and their typical use
- 4:12cases to match technique to utility. This ensures you are
- 4:15selecting methods based on established best practices
- 4:17rather than guesswork. Finally, unrealistic expectations often
- 4:20derail projects when teams deviate from the defined
- 4:22performance profile or project constraints. Stick strictly to
- 4:25the defined performance profile and project constraints,
- 4:27including budget and timeline limits. This ensures candidates
- 4:30and team members are prepared for the specific research
- 4:32demands, keeping the project on track and effective.
- 4:37Welcome to The Take. I'm joined by a fellow practitioner and
- 4:40together we're going to dig into Choosing Research Techniques: A
- 4:43Practical Guide — what it means in practice and what's worth
- 4:46remembering.
- 4:48[HOST] That four-step framework feels solid, but I wonder if the
- 4:51risk assessment step gets skipped when deadlines loom.
- 4:54[LEARNER] It absolutely does. We usually jump straight to
- 4:57recruiting because we’re terrified of missing the sprint
- 5:00goal. [HOST] And that’s where the unvalidated model pitfall
- 5:03hits us hardest. We assume we know the users without verifying
- 5:06with customer service data. [LEARNER] True. I’ve seen teams
- 5:09build prototypes based on internal hunches that completely
- 5:12missed the mark. [HOST] The lesson suggests verifying those
- 5:15models early to mitigate that specific danger. It sounds
- 5:18simple, but it’s often the hardest sell to stakeholders.
- 5:21[LEARNER] They want speed, not verification. But here’s where
- 5:24AI might actually help us bridge that gap quickly. [HOST] You
- 5:27mean using generative tools to simulate user interviews or
- 5:30analyze support tickets? [LEARNER] Exactly. We could feed
- 5:33recent support logs into an LLM to spot high-risk areas faster
- 5:36than manual review. [HOST] That’s a clever workaround for
- 5:39the time constraint. It gives you data-backed insights without
- 5:43the full logistical overhead. [LEARNER] Just remember it’s a
- 5:46proxy. It doesn’t replace the nuance of actual contextual
- 5:49inquiry. [HOST] Right. It helps you define the scope, but you
- 5:52still need real humans for the validation phase. [LEARNER] So
- 5:55we use AI to sharpen the questions, then traditional
- 5:58methods to answer them. [HOST] Precisely. We keep the human in
- 6:01the loop for the critical decisions, even if the prep work
- 6:04is automated.