 Home
 Studying in New Zealand
 Qualifications and standards
 NCEA
 Māori & Pasifika

Providers and partners
 About education organisations
 NZQA's quality assurance system for tertiary education organisations
 Quick links to NZQF documents
 Approval, accreditation and registration
 Consistency of graduate outcomes
 Monitoring and Assessment
 Selfassessment
 External evaluation and review
 Assessment and moderation of standards
 Submitting results and awarding qualifications
 The Education (Pastoral Care of International Students) Code of Practice
 Offshore use of qualifications and programmes
 Guidelines and forms
 About us
Assessment Report
Level 3 Statistics 2017
Standards 91584 91585 91586
Part A: Commentary
The 2017 papers gave a similar overall distribution of grades as previous papers with the exception of AS9154 (Statistical Reports), where the Excellence rate was considerably higher than in recent years. Candidates who performed well across the standards were able to integrate statistical and probability concepts, calculations and statistical reasoning. This included using contextual information to support their answers. Successful candidates demonstrated strong skills in working with confidence intervals, probability theory and probability distributions, as well as sound reasoning with data and models. Weaker candidates often did not supply numerical evidence to support their responses, or only supplied numerical evidence with no supporting explanations.
Part B: Report on standards
91584: Evaluate statistically based reports
Candidates who were awarded Achievement commonly:
 identified and described explanatory and response variables
 used specific skills such as calculating margin of error
 calculated confidence intervals but could not interpret them in context
 read numbers from graphs correctly or identified some key aspects of a report
 understood as sample size increases, the margin of error decreases
 could recall names of nonsampling errors
 had limited understanding of different study designs
 used ‘common sense’ to comment on key features of the report.
Candidates who were assessed as Not Achieved commonly:
 did not use information from the provided statistical reports
 did not attempt many question parts
 gave generic or rote learnt answers that were not linked to the report
 did not understand the different types of confidence intervals
 could not read a simple graphical display or interpret text for statistical values
 identified a key feature but failed to describe its relevance to the context.
 failed to identify the need for a margin of error in statistical survey reports
 could not calculate a margin of error
 lacked familiarity with experimental design principles, interpreting statistical inference, interpreting graphs, critiquing causal relationships, and interpreting margins of error.
Candidates who were awarded Achievement with Merit commonly:
 used specific quotes from statistical reports to support their answers
 understood how to make a call using a confidence interval about claims made in context
 demonstrated understanding of the 30%70% guidelines for using the rule of thumb margin of error
 described and understood different study designs with some interpretation
 identified and described a relevant nonsampling error in context
 presented an argument that used some statistical reasoning and evidence from the statistical report
 demonstrated understanding of study designs, specifically the use of random allocation of treatments for an experiment.
Candidates who were awarded Achievement with Excellence commonly:
 focused on the specific claims being made and whether the statistical processes described in the statistical report supported the claim
 quoted directly from the report as evidence and used this evidence combined with statistical and contextual knowledge to construct a convincing argument
 responded to a claim concerning the difference between two percentages by constructing and interpreting a confidence interval, and then explaining that a claim could not be made because 0 was included in the interval, or because the values of the confidence interval were either both positive or both negative
 were succinct in their responses and presented clear, wellsupported discussion points rather than lengthy generic statements or descriptions
 integrated statistical and contextual knowledge in their responses when discussing issues with a statistical process or claim presented in the statistical report.
Standard specific comments
Every statistical report will have aspects that candidates will not get to discuss, due to time constraints and the need for a balance of questions. Candidates often tried to use nonrelevant parts of the statistical report to answer a question that was not asked. For example, in question one, many candidates wanted to discuss examples of confounding variables despite not being specifically asked for this.
Candidates need to look carefully at the question that was asked, including what quotes or headlines from the article they are being asked to discuss. Specific evidence from the report needs to be used to support any arguments or points made. For example, many candidates discussed the use of online surveys for Q2(c)(ii) in general terms, writing what were previously learned responses, without linking to anything specific from the statistical report or explaining why the use of online surveys might cause bias (linking back to the target population and what was being measured through the survey).
Many candidates either did not interpret the confidence intervals constructed in words, or incorrectly combined their interpretation with “making a call” as one sentence.
91585: Apply probability concepts in solving problems
Candidates who were awarded Achievement commonly:
 understood concepts, but didn’t supply numerical evidence
 were able to draw and understand a twoway table
 were able to find values in tables and charts
 identified relevant probability concepts and contextual information from the problem.
Candidates who were assessed as Not Achieved commonly:
 did not use calculations anywhere to support statements
 often gave proportions greater than one
 mixed up the tests for independence and mutually exclusive
 gave answers only
 struggled to relate to or understand the context of the question
 did not show an understanding of the difference between experimental and theoretical probability
 could not calculate conditional probabilities on a tree diagram
 were expecting to see a relative risk question and used it when not needed
 confused sample size with sampling variability.
Candidates who were awarded Achievement with Merit commonly:
 were able to link their ideas with the relevant probability concept, for example when proving events were not independent nor mutually exclusive
 were able to find the conditional probabilities on the tree diagram
 were able to support their ideas with relevant calculations
 were able to understand data given in different forms
 had clear working and clear answers
 used probability notation correctly
 could construct a 3event tree diagram.
Candidates who were awarded Achievement with Excellence commonly:
 could correctly construct a 3event Venn diagram and use it to solve a problem
 were able to calculate appropriate theoretical probabilities for a game involving probability
 showed an understanding of the context when interpreting their answers
 were able to develop a logical chain of reasoning using probability theory to deduce a relevant probability.
Standard specific comments
Candidates showed some lack of understanding about sampling variability and sample sizes and how this affects probabilities.Many candidates did not understand the entries on the second and subsequent columns on the tree diagram are conditional.A very few candidates spent much time drawing huge tree diagrams for Q3b when it was not required because it was a symmetrical distribution. Most candidates did not realise that they had to back up the independence and mutually exclusive proofs with numerical evidence.
91586: Apply probability distributions in solving problems
Candidates who were awarded Achievement commonly:
 selected appropriate probability distribution models
 calculated simple probabilities using probability distribution models
 understood factors that should be kept constant, when modelling practical situations
 understood terms such as ‘less than’, ‘at least’, ‘more than’, ‘at most’
 calculated the expected value (mean) of a discrete random variable
 calculated the variance of a discrete random variable
 used discrete random variables represented as probability distribution tables
 identified when a particular model may be appropriate or inappropriate to use for experimental data, without clearly justifying and linking limitations of the model to the context.
Candidates who were assessed as Not Achieved commonly:
 used incorrect methods, e.g. chose the wrong model or performed incorrect probability calculations
 confused the inequalities and hence, incorrectly calculated probabilities.
 misinterpreted inequalities expressed as text
 could not correctly calculate means, variances and probabilities from a probability distribution table
 were unable to compare features of an experimental distribution with a probability distribution graph or table
 made general statements unsupported by evidence
 did not complete sufficient questions.
Candidates who were awarded Achievement with Merit commonly:
 linked features of the probability distribution model to the context, e.g. identified appropriate limitations of the model and explained them in terms of the context or gave conditions for a probability distribution in context
 completed multistep problems across a range of distributions; could link their responses to the stated context and answer the questions asked
 communicated their thinking using appropriate statements, e.g. stating probability distribution model and parameters, correctly using probability notation and providing evidence or calculations for any general statements they made.
Candidates who were awarded Achievement with Excellence commonly:
 devised a strategy to solve multistep probability distribution problems
 discussed the appropriateness (or inappropriateness) of a model by considering features of the probability distribution, statistical evidence and/or the context
 compared features of an experimental distribution and a probability distribution model and linked these features to the context of the experimental distribution
 clearly explained and justified with evidence their reasoning.
Standard specific comments
The paper identified candidates’ confusion/ misconceptions about a number of concepts, particularly; randomness, independence, symmetry and skew, continuous and discrete variables, what a mean represents, the relationship of a mean to variance and standard deviation, and, in the Poisson distribution, what the idea of proportionality means in terms of a rate in a particular context.
When discussing conditions and features of all probability distribution models, many candidates did not consider each condition in terms of the given context. Candidates need to be able to explain how the conditions and/or features of a probability distribution model (e.g. randomness, independence, skew, bellshaped, upper/lower bounds) are appropriate or inappropriate to the given context.
Given a clear graph displaying the experimental and theoretical proportions for a situation, candidates should use this visual information to construct an argument. Comments should use the graph to compare the theoretical with experimental distributions. Comments need to be supported by specific numerical evidence in context. Many candidates did not support their statements with statistical and/or numerical evidence.
Mathematics and Statistics subject page
Previous years' reports
2016 (PDF, 228KB)