BUSINFO700 Analysis of Business Problems
The Complete Study & Assessment Guide · 2026BUSINFO700 Overview
- Master of Business Analytics
- Postgraduate
- Analyses run in R
- Test on 4 November 2026
- Ten teaching weeks in 2026
Analysis of Business Problems is a postgraduate course in the University of Auckland's Master of Business Analytics.
- Assessed by An individual email assignment, an invigilated closed-book test and a group presentation.
- Hardest step Choosing the analysis from the measurement level of each variable, then explaining the output in business terms.
- How to prepare Time yourself on practice questions and justify every DELTA-Plus score with a fact from the case.
How BUSINFO700 is assessed
| Component | Weight | Format |
|---|---|---|
| Email assignment | 20% | Individual: DELTA-Plus report with AI evaluation, an email to the CEO and a reflection; due Friday 16 October 2026, 23:59 |
| Test | 50% | Week 6, Wednesday 4 November 2026, 6:30 pm to 8:30 pm; in-person, invigilated Canvas Quiz; Closed-book Test on Weeks 1 to 6 |
| Group presentations | 30% | Group task introduced in Week 7; presentations scheduled for Week 10, from Monday 30 November 2026 |
The three weights come from the course schedule and total 100%; the email assignment page and the test instructions repeat their own 20% and 50%. The schedule table on the Week 7 lecture slides lists a hand-in assignment at 30% in Week 8 and presentations at 20%, so confirm the current group-task split on Canvas. No hurdle is stated for any component; check the course outline on Canvas if you need certainty.
Current BUSINFO700 dates
| Date | Item | Control |
|---|---|---|
| October 16, 2026 | Email assignment | Due Friday by 23:59. |
| November 4, 2026 | Test | 6:30 pm to 8:30 pm; enter your allocated room from 6:15 pm. |
| November 30, 2026 | Group presentations | Week 10 session; room to be announced. |
Dates are as published in the 2026 course schedule and test instructions on Canvas. Confirm exact deadlines and submission settings in the live LMS.
What BUSINFO700 covers
The course runs on two tracks. Chapters 1 to 4 build the organisational track: framing a business problem, scoring maturity with DELTA-Plus, writing to a chief executive and matching machine learning to a maturity stage. Chapters 5 to 9 build the statistical track in R, from describing one variable to testing relationships, segmenting customers and estimating drivers with regression.
Business Problems, Big Data and the R Workflow
Course aims, learning outcomes, why R, data types, vectors and data frames, and a reproducible script workflowThis chapter sets up the habit the whole course rewards: begin with the business decision, identify the evidence that could change it, and only then choose an analysis.
- A managerial view of small and big data
- The six course learning outcomes
- Why the labs use R and RStudio
- Vectors, factors, data frames and lists
- Missing values and silent errors
- Describing data before modelling it
The DELTA-Plus Model and Analytics Maturity
Seven success factors, five maturity stages, scoring a case from evidence and choosing one recommendationThis chapter teaches the model that anchors both the email assignment and the largest question on the practice test.
- Where DELTA and the five stages come from
- The seven success factors and their evidence
- Placing each factor on a maturity stage
- Uneven profiles and targeted investment
- Three pitfalls when recommending
- The timed DELTA-Plus test question
Writing the Maturity Email and Judging AI Output
The three-part email assignment, its rubric criteria, prompt iteration and turning scores into manager languageThis chapter covers the individual email assignment, worth 20% of the grade and due on Friday 16 October 2026. The task had three parts with different readers: a report for the lecturer comparing your own DELTA-Plus analysis with the output of two generative AI programs, an email to the chief executive describing the company's maturity in words and giving exactly three urgent recommendations, and a reflection on your judgement and the limits of AI.
- The three parts and who reads each
- Word ranges and format requirements
- The six rubric criteria and their points
- Doing your own analysis before using AI
- Writing maturity in manager language
- Choosing exactly three urgent recommendations
Machine Learning Across the Analytics Stages
Descriptive to automated analytics, supervised and unsupervised learning, and the questions to ask before modellingThis chapter connects the maturity stages from DELTA-Plus to the analysis a team can realistically run, following the course reading by Paas (2023).
- Five stages mapped to analytical methods
- Inductive applied analytics versus deductive research
- Rule-based segments and dashboards
- Unsupervised learning for customers and products
- Supervised learning, training and test sets
- Judging readiness for a predictive model
Univariate Analysis: Variables, Centre and Spread
Measurement levels, frequency tables, mean, median and mode, variance and standard deviation, skewness and z-scoresUnivariate analysis describes one variable at a time, and this chapter follows the Week 3 lecture and lab in starting with the question that controls everything else: is the variable non-metric (binary, nominal or ordinal) or metric (interval or ratio)?
- Non-metric and metric measurement levels
- Frequency tables with percent and cumulative percent
- Mean, median and mode on skewed data
- Variance and standard deviation with n minus 1
- Skewness and the slight and strong cut-offs
- Z-scores for comparing different scales
Hypothesis Testing, Causality and Error Types
Conditions for causality, confounding, null and alternative hypotheses, one- and two-tailed tests, Type I and Type II errorsThis chapter covers the reasoning that sits underneath every bivariate test in the course.
- Four conditions for causality
- Confounding and systematic error in large datasets
- Null and alternative hypotheses
- One-tailed versus two-tailed tests
- Type I and Type II errors
- Writing a conclusion a manager can use
Choosing a Bivariate Test: Chi-Square, T-Test, ANOVA
Matching measurement levels to a test, expected counts, reading p-values, ANOVA with Levene and post-hoc testsThis chapter turns test selection into a routine you can run in seconds.
- Classifying variables as metric or non-metric
- Cross-tables, expected counts and chi-square
- Fisher's exact test for small samples
- Two-group comparisons with the t-test
- ANOVA, Levene and post-hoc tests
- Writing the manager's verdict from output
Cluster Analysis for Customer Segmentation
Segmentation logic, Euclidean distance, standardising, k-means, silhouette scores and profiling clustersThis chapter shows how k-means cluster analysis turns a varied customer base into a few segments a business can act on.
- Cut-off segmentation versus data-driven segments
- Euclidean distance and nearest-centre assignment
- Standardising variables with scale()
- Choosing k with silhouette scores
- Analyst judgement on statistical, visual and business grounds
- Profiling and naming clusters
Linear Regression for Business Drivers
The six-step regression procedure, beta coefficients, R-squared and the F-test, assumptions, multicollinearity and predictionLinear regression is the course's first supervised learning method and its main tool for estimating business drivers. It is taught in Week 8, after the 4 November test, so it matters for later course work rather than for that test.
- Least squares, intercept and slope
- The six steps of a regression analysis
- Interpreting betas, including a 0/1 dummy
- Ranking drivers by t-statistic
- R-squared, adjusted R-squared and the F-test
- Assumptions, diagnostic plots and multicollinearity
Its official description promises a managerial perspective on using small and big data to identify problems, analyse them and make decisions, always with an eye on the commercial setting that analytics teams serve and on how departments collide when data start to drive decisions. In practice the course runs on two tracks that meet in every assessment. The first track is organisational.
Davenport's DELTA-Plus model scores a company on its data, enterprise orientation, leadership, targets, analysts, technology and analytical techniques, and places it on a five-stage ladder that runs from analytically impaired to analytical competitor. The second track is statistical and is taught in R.
Week 3 covers frequency tables, measures of centre and spread, skewness and z-scores; Week 4 adds causality, hypothesis testing, the chi-square test and the t-test; Week 6 continues with ANOVA and correlation; cluster analysis and linear regression follow in Weeks 7 and 8. A reading by Paas (2023) joins the two tracks by showing which analyses and which kinds of machine learning belong at each maturity stage.
The email assignment asks you to assess a New Zealand retailer case with DELTA-Plus, critique the output of two generative AI tools and write a recommendation email to the chief executive.
The invigilated test on 4 November asks for the same judgement under time pressure: apply DELTA-Plus to a short case, select the right analysis for a business question, and interpret R output, including which error type the analyst could have made. This guide follows that order, gives every method an original worked example, and closes with a timed practice set and a test-night checklist.
Worked example · free
Compute a chi-square test from a two-by-two table
- 1Both variables are non-metric (plan type and joined yes or no), so the chi-square test of association applies. H0: plan type and class joining are not associated; H1: they are associated.
- 2Expected count = row total × column total / 200. With 70 joiners overall, monthly expects 120 × 70 / 200 = 42 joiners and 78 non-joiners; annual expects 28 joiners and 52 non-joiners. Every expected count is at least 5.
- 2Chi-square = (30 − 42)²/42 + (90 − 78)²/78 + (40 − 28)²/28 + (40 − 52)²/52 = 3.43 + 1.85 + 5.14 + 2.77 = 13.19 with 1 degree of freedom.
- 113.19 exceeds the 0.05 critical value of 3.84 (p is about 0.0003), so reject H0: annual members joined classes at 50% against 25% for monthly members.
Key terms
- DELTA-Plus model
- Davenport's framework that rates an organisation on data, enterprise, leadership, targets, analysts, technology and analytical techniques.
- Analytics maturity stage
- One of five levels, from analytically impaired to analytical competitor, describing how far analytics drives an organisation.
- Metric variable
- A variable whose numeric differences carry quantitative meaning, measured on an interval or ratio scale.
- Non-metric variable
- A categorical variable, binary, nominal or ordinal, whose codes do not measure quantity.
- Standard deviation
- The square root of the variance, giving typical distance from the mean in the variable's own units.
- Skewness
- A measure of asymmetry: positive values indicate a long right tail, negative values a long left tail.
- Z-score
- The number of standard deviations an observation lies above or below its variable's mean.
- Null hypothesis
- The default claim of no difference or no association that a significance test tries to reject.
- Type I error
- Concluding from a sample that an effect exists when it does not exist in the population.
- Type II error
- Failing to detect, in a sample, an effect that does exist in the population.
- Chi-square test
- A significance test comparing observed and expected counts to judge whether two categorical variables are associated.
- One-way ANOVA
- A test of whether a metric variable's mean differs across three or more groups.
- K-means clustering
- An unsupervised method that assigns each observation to the nearest of k cluster centres.
- Coefficient of determination
- R-squared: the share of variation in the dependent variable that a regression model explains.
BUSINFO700 FAQ
How is Analysis of Business Problems assessed?
Three components make up the grade. The individual email assignment is worth 20% and was due on Friday 16 October. The invigilated test is worth 50% and runs on 4 November from 6:30 pm. Group presentations make up the remaining 30% and are scheduled for Week 10. Check Canvas for the current group-task details, because one lecture slide shows an older split.
What does the November test cover?
The scope is the lecture and lab material from Weeks 1 to 6, including the Monday 2 November session on bivariate statistics. The instructions say the question types match the practice test but the cases are longer and some questions have more parts. The paper totals 100 points, and each question shows its points and a suggested time.
Can I use notes, software or AI during the test?
No. It is a closed-book test sat on a lab PC in an allocated room, with answers typed into the Canvas quiz. Only the Canvas test pages may be open. Personal devices, notes, spreadsheets, translators, dictionaries, browsers and generative AI tools are all excluded, although pens and highlighters are allowed.
What is the DELTA-Plus model used for?
It assesses an organisation's analytics maturity. Davenport's original five elements are data, enterprise, leadership, targets and analysts, and the 2018 primer adds technology and analytical techniques. Each factor is placed on one of five stages, so a case analysis scores the factors from evidence, identifies the weakest ones and recommends where to invest first.
How do I decide between chi-square, a t-test and ANOVA?
Start from the measurement level of both variables. Two non-metric variables call for a cross-table and chi-square. A binary grouping variable with a metric outcome calls for a t-test comparing two means. A grouping variable with three or more categories and a metric outcome calls for ANOVA, followed by post-hoc tests to locate which groups differ.
Why spend weeks on basic statistics in an analytics degree?
The labs argue that descriptive and bivariate statistics show what is happening in the data before any complex model is trusted. They expose coding errors such as impossible ages, outliers that distort averages and confounded relationships. One lab shows a consultant's churn model whose surprising age effect needs exactly this kind of bivariate check.
What happens if an assignment is submitted late?
The programme policy states that late assignments usually incur a minimum deduction of 5% per day. Extensions are possible only in exceptional circumstances, such as illness, and are decided case by case by the lecturer, who may ask for evidence like a medical certificate. Apply well before the due date rather than afterwards.
Questions students ask in each BUSINFO700 chapter
- Do I need prior programming experience for the R labs?
- The first lab starts from installing R and RStudio and builds up from the console, the four RStudio panes and the working directory to vectors and data frames. Code is scaffolded step by step, and the notes point to the free online book R for Data Science and to each function's help page. Chapter 01 · Business Problems, Big Data and the R Workflow
- What does each letter in DELTA stand for?
- D is data, E is enterprise, L is leadership, T is targets and A is analysts. These five elements come from Analytics at Work (2010). The 2017 revision added technology and analytical techniques, called the Plus factors, giving seven elements that are each placed on one of five maturity stages. Chapter 02 · The DELTA-Plus Model and Analytics Maturity
- Could I use ChatGPT for the email assignment?
- Yes, within rules. Your own DELTA-Plus analysis had to come first, then two generative AI programs were applied and compared with it. AI could help draft or polish the email, but you remained responsible for checking the recommendations, citing AI in APA 7th style and placing every prompt and output in the appendix. Chapter 03 · Writing the Maturity Email and Judging AI Output
- What is the difference between supervised and unsupervised learning?
- Unsupervised learning has no outcome variable; it finds structure, such as customer clusters or products bought together. Supervised learning predicts a recorded behaviour, such as responding to an offer or churning, from explanatory variables. Ask whether there is a target behaviour, and the branch follows. Chapter 04 · Machine Learning Across the Analytics Stages
- Why does the course insist on knowing the measurement level first?
- The level determines which summaries make sense. A mean of postcodes or of income band codes is arithmetic on labels and tells you nothing. Non-metric variables get frequency tables, while metric variables get the mean, median, mode, standard deviation and skewness. The same rule later decides which bivariate test applies. Chapter 05 · Univariate Analysis: Variables, Centre and Spread
- When should I use a one-tailed test instead of a two-tailed one?
- Only when well-established theory predicts the direction of the effect before you see the data. In business intelligence work theories are usually weak and the opposite effect is often possible, so the Week 4 lecture recommends two-tailed tests as the default. A one-tailed test would leave no room to detect the unexpected direction. Chapter 06 · Hypothesis Testing, Causality and Error Types
- When should I use Fisher's exact test instead of chi-square?
- Use it when the sample is small or expected cell counts drop below 5, because the chi-square approximation then becomes unreliable. In R, check the expected table first; if several cells are under 5, run fisher.test on the same two variables and report its p-value instead. Chapter 07 · Choosing a Bivariate Test: Chi-Square, T-Test, ANOVA
- Why must variables be standardised before k-means?
- k-means compares customers by Euclidean distance, so a variable measured on a wide range, such as annual spend in dollars, would swamp one on a narrow range, such as visits per month. Converting each variable to z-scores with scale() gives every variable an equal say in the distance. Chapter 08 · Cluster Analysis for Customer Segmentation
- What are the six steps of a regression analysis in this course?
- Formulate the null and alternative hypotheses for each driver, collect data, generate the output, interpret it, assess model fit and check that the assumptions hold. Interpretation covers which drivers are significant, their ranking and the meaning of each beta; model fit covers R-squared, adjusted R-squared and the F-test; the last step includes multicollinearity. Chapter 09 · Linear Regression for Business Drivers
How to prepare for the assessments
Work the two tracks in parallel rather than one after the other. For the organisational track, take any short company description, list the facts that bear on each DELTA-Plus factor, and give each factor a score with one sentence of evidence before choosing a single recommendation. Then repeat the exercise without the stage table in front of you, because the test is closed-book.
For the statistical track, build a one-page decision sheet in your own words that maps the measurement level of each variable to its analysis: frequency table, mean and standard deviation, chi-square, t-test, ANOVA or correlation. Rehearse it on short business scenarios until the choice takes seconds.
Next, practise reading R output aloud: name the test, state the null hypothesis, quote the statistic and the p-value, give the business conclusion, and add one sentence on bias, missing values or causality. Finish each session with one timed question from the practice set at the suggested time, then compare your answer with the marking points in that chapter.
In the final week, sit the full practice set at its suggested times and note which question type takes longest.
Your AI Statistics tutor for BUSINFO700
Stuck on a hard BUSINFO700 question? Sia is AskSia’s AI Statistics tutor — ask any BUSINFO700 Analysis of Business Problems question and get a clear, step-by-step explanation grounded in how the course is actually taught and assessed. Read this whole study guide free, then take your hardest questions to Sia.
