41297 Chap.3 Defining a Business Problem and Research Questions
Defining a Business Problem and Research Questions
Framing before calculating
Lecture 2 makes problem definition the first analytical move.
A broad business interest is narrowed into analytical questions that name variables, then restated as objectives that open with an action such as determine, estimate, identify or predict.
Examples from many business areas
The lecture's examples range across sport, healthcare, retail, marketing, human resources, finance, banking, tourism, education, transport, real estate and the adoption of artificial intelligence, so you can model your own wording on a problem close to your interests.
The seminar's two forms
Seminar 1 turns the framing into paperwork.
The topic selection form records your dataset, two reasons for choosing it, its quality and two or three research questions. The data description form records its size and every variable, and the dataset needs over 200 observations and five or more relevant variables.
What this chapter covers
- 01
Defining and structuring a business problem
- 02
Analytical questions by business topic
- 03
Restating problems as objectives
- 04
The topic selection form
- 05
The data description form and cropped datasets
- 06
The 200-observation and five-variable rule
Worked example · free
Frame a delivery problem and run the dataset check
- 1Research question: which order and driver characteristics are associated with more minutes late?
- 1Objective: identify the factors linked to late delivery to guide slot planning.
- 1Observations: 1,500 orders is more than the 200 required.
- 1Relevant variables: slot, distance, items, driver experience and minutes late make five.
- 1Exclude the order ID, which answers no question, so the file qualifies with exactly five.
Key terms
- Research question
- A precisely worded question, ending in a question mark, that names the variables a dataset must contain to answer it.
- Research objective
- A statement opening with an action verb that says what an analysis will determine, estimate, identify or predict.
- Cropped dataset
- The subset of rows and columns actually selected for analysis from a larger original file.
Defining a Business Problem and Research Questions FAQ
How many research questions should I write in the first seminar?
The topic selection task asks for two or three research questions, objectives or challenges, and the form has space for three. Fewer than two leaves too little for the later methods to answer; more than three usually means the topic is still too broad and needs narrowing.
What is the difference between a research question and an objective?
A research question asks something, ends in a question mark and names its variables, such as whether training hours relate to productivity. An objective states what the analysis will deliver and opens with an action verb, for example estimate the effect of training hours on output per worker.
Does my dataset have to come from the lecture's list of sources?
No. Seminar 1 lets you pick from the repository links in the lecture or follow your own interest. Whatever the origin, the dataset still needs justifying with facts and figures, a quality judgement, and enough rows and relevant variables to pass the size rule.
Why record both the original and the cropped dataset sizes?
The data description form separates the file you downloaded from the rows and columns you will actually analyse. Showing both lets a reader see what you removed, and it is the cropped version whose variables you classify by type, level, unit and time.
Assessment move
Draft three versions of your topic in one sitting: a broad interest, a research question that names variables, and an objective that starts with a verb. Read the lecture's example lists for a business area near yours and borrow the structure, not the wording. Then open the dataset and tick each named variable against a column; if any is missing, rewrite the question rather than forcing the data.
Finish by counting rows and relevant variables against the seminar rule and filling both forms before Seminar 2.
Working through Defining a Business Problem and Research Questions in 41297? Sia is AskSia’s AI Statistics tutor — ask any 41297 Defining a Business Problem and Research Questions question and get a clear, step-by-step explanation grounded in how 41297 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.