ASSESSMENT EXEMPLAR WITH GUIDE FOR INSTRUCTORS: STATISTICAL INDEPENDENCE Topic: ☐Microeconomics ☐Macroeconomics ☒Statistics/Econometrics Level: ☐Bachelors ☐Masters ☒Both/Either Core concepts: Statistical independence; test statistics and their distributions Learning domains: ☒Knowledge ☒Application ☒Data analysis ☒Communication ☒Reflection Type(s) of task (can be more than one) ☒Constructed response given stimulus material (clip/narrative/etc.) ☐Constructed response with no stimulus material ☒Quantitative analysis ☐Independent data sourcing ☐Independent research (not data sourcing) ☒Written presentation ☒Oral presentation ☐Group collaboration ☐Use of nominated maths/stats software (specify): ☐Use of other technology (specify): ☐Other (specify): ________________________________________ The core statistical concept underpinning this assessment item is independence, framed in an examination of the quality of the statistical relationship between categorical variables. Students must apply abstract concepts related to bivariate relationships, such as marginal distributions and statistical independence, to an actual data context, and use statistical testing (including applying formulas) and reasoning to understand what is happening in that particular context. The item asks students to analyse data displayed in a cross-tabulation and to express their interpretation of the results of statistical tests. Students are also asked to provide reasons why particular results may have been found. Situating students’ results within the problem of education policy and the debate about academic standards then gives rise to some more reflection-oriented components, and a clear direction to extend the item for Masters students. The callouts to the learning standards assessed by different components of the items appear in square brackets within the items, and should be removed before the items are used. 1 Bachelors level task: You have been contracted by the Head of the School of Basket-Weaving to perform some statistical analysis in order to guide him in reviewing the School’s policies related to teaching and learning. You are provided with a data set capturing individual students’ grades in a course of 250 students that used two different types of tutorials: Type A and Type B. The table below provides a summary of this information. Table 1: Basket-Weaving Students’ Results by Type of Tutorial Fail Pass Credit Distinction High Distinction Type A 13 59 26 6 1 Type B 12 71 39 19 4 TOTAL 25 130 65 25 5 i. ii. iii. iv. v. vi. vii. TOTAL 105 145 250 What is the probability that a student chosen randomly from the class was allocated into a Type {TYPE} tutorial and received at least a {GRADE} grade? [KB1,DB2] Suppose that all {GRADE}-earning students are collected into one spreadsheet. What is the probability that a student chosen at random from that spreadsheet was enrolled in a Type {TYPE} tutorial? [KB1,DB2] Are the events “being allocated into a Type B tutorial” and “receiving a Distinction grade” independent events? Provide a justification for your answer. [KB1,DB2,CB1] Suppose a formal test of the independence of tutorial type and grade is performed using these data, and the null hypothesis of independence is rejected at the 5% level of significance. (You do not have to do this test.) Give two possible reasons why the test might have indicated a lack of independence. [RB1,CB1] The Head of the School of Basket-Weaving is considering instituting a policy that all classes in the School be graded according to a common grade distribution given by: 3% HD, 12% D, 25% C, 50% P, 10% F Test whether the actual overall distribution of grades for the class you are analysing complies with this distribution. Use a significance level of 1%. Be sure to state clearly your hypothesis, decision rule and conclusion. [KB1,DB1, DB2] What conclusions can you draw from your analysis about the potential effects of instituting a common-grade-distribution policy across the entire School? [AB2,CB1] What other information would you like to have in order to generate advice about the likely School-wide effects of instituting a common-grade-distribution policy? What would you do with this additional information in order to generate this advice? [AB2,RB2,CB1] 2 Masters level task: You have been contracted by the Head of the School of Basket-Weaving to perform some statistical analysis in order to guide him in reviewing the School’s policies related to teaching and learning. You are provided with a data set capturing individual students’ grades in a course of 250 students that used two different types of tutorials: Type A and Type B. The table below provides a summary of this information. Table 1: Basket-Weaving Students’ Results by Type of Tutorial Fail Pass Credit Distinction High Distinction Type A 13 59 26 6 1 Type B 12 71 39 19 4 TOTAL 25 130 65 25 5 i. ii. iii. iv. v. TOTAL 105 145 250 Are the events “being allocated into a Type B tutorial” and “receiving a Distinction grade” independent events? Provide a statistical justification for your answer, and interpret what your answer could mean in the context of the School’s teaching and learning practice in this course. [KM1,DM2] Suppose a formal test of the independence of tutorial type and grade is performed using these data, and the null hypothesis of independence is rejected at the 5% level of significance. (You do not have to do this test.) Give two possible reasons why the test might have indicated a lack of independence. [RM1,CM1] The Head of the School of Basket-Weaving is considering instituting a policy that all classes in the School be graded according to a common grade distribution given by: 3% HD, 12% D, 25% C, 50% P, 10% F Test whether the actual overall distribution of grades for the class you are analysing complies with this distribution. Use a significance level of 1%. Be sure to state clearly your hypothesis, decision rule and conclusion. What do your results imply about the effects of implementing a similar common-grade-distribution policy across the entire School? [KM1,DM1,DM2,CM1] What other information would you like to have in order to generate advice about the likely School-wide effects of instituting a common-grade-distribution policy? Why? What would you do with this additional information in order to generate this advice? [AM2,RM1,RM2,CM1] In light of your knowledge of the Australian higher education sector, what costs and benefits for students and for the university would you predict to arise in the long run if such a common-grade-distribution policy were implemented across the sector? Are there any ethical concerns with the implementation of this policy that you would point out as an economist? Why or why not? [RM3,CM1] 3 Advice to assessors: The type of communication used to express the answers to parts (vi) and (vii) in the Bachelors item and parts (iii), (iv), and (v) in the Masters item can be written, which is the default, but could also be oral. Masters students could use part (v) at the basis for a debate about Australian higher education policy, possibly incorporating a collaborative component if debate preparation occurs within teams, or a role-playing game where different students play the role of The Statistician, The Head of School, the Vice-Chancellor, The Teacher, and The Student. Masters students could also be asked to prepare reports on this same question for different audiences: one for consumption by the Head of School or the Vice-Chancellor, and one for consumption by the Commonwealth Department of Education, for example. Masters students may be given reading material about the higher education sector in order to respond knowledgably to part (v). Such material should be current and drawn from ongoing debates on issues relating to academic standards (e.g., op-eds from The Australian or articles from The Conversation), and could also include historical reports or articles about grade distributions in higher education over time. For a further extension, they could be tasked with finding data sources that could be used as input into statistical analyses that would help inform policy-setting in this area. The question frame can be used as a graft for other questions that query students’ skills in evaluating bivariate relations for purposes of education policy-setting – for example, an examination of the student-level association between grades and lecture attendance, tutor, department, program of study, or other discrete factors. 4 Learning domain Learning outcomes Bachelor Degree Masters Degree Knowledge KB1 Bachelor graduates will be able to identify, coherently explain and synthesise core economic concepts KM1 Masters graduates will be able to identify, coherently explain and synthesise core and advanced economic concepts, including recent developments in the discipline Application Bachelor graduates will be able to: Masters graduates will be able to: AB1• frame problems in terms of core economic concepts and principles AM1• frame and critically analyse problems in terms of core and advanced economic concepts and principles AB2• apply economic reasoning and analytical skills, in order to make informed judgments and decisions AM2• apply advanced economic reasoning and analytical skills, including quantitative techniques where appropriate, in order to make informed judgments and decisions AM3• plan and execute a research-based project Data analysis Bachelor graduates will be: Masters graduates will be able to: DB1• able to use economic data to address typical problems faced by economists DM1• select and apply an appropriate empirical method to address typical problems faced by economists DB2• aware of, and able to implement, basic empirical techniques and interpret the results DM2• critically evaluate the results Communication CB1 Bachelor graduates will be able to present a clear and coherent exposition of economic knowledge, ideas and empirical evidence both orally and in writing, individually or in collaborative contexts CM1 Masters graduates will be able to communicate complex ideas clearly and coherently, in written form and interactive oral form to expert and non-expert audiences, individually or in collaborative contexts Reflection Bachelor graduates will be able to reflect on: Masters graduates will be able to reflect on and evaluate: RB1• the nature and implications of assumptions and value judgments in economic analysis and policy RM1• the nature and implications of assumptions and value judgments in economic analysis and policy RB2• interactions between economic thinking and economic events, both historical and contemporary RM2• interactions between economic thinking and economic events, both historical and contemporary RB3• the responsibilities of economists and their role in society RM3• the responsibilities of economists and their role in society 5