MKTG3014 Assessment Task 2 – Discrete Choice Experiment
Task Description: This is an individual assessment that requires students to design a Choice-based Conjoint (CBC) study and a Rating-based Conjoint study, and complete an appropriate set of analyses on the data.
Type of Collaboration: Individual
Length: Maximum 2200 words (excluding Cover-page, Headings, Sub-headings, References, and Appendices)
Submission: Submit your report through the Turnitin link on iLearn, no hard copies of soft copies via email will be accepted. You must submit by the due date
as specified
Due: Friday, 4 November 2022, 11.55pm (Week 13)
Weighting: 40%
Instructions
Part I
Students should design and conduct a Choice-Based Conjoint (CBC) experiment in the Sawtooth software. Students would select a product (or service) of interest and identify a minimum of 3 key attributes of the product (clear theoretical and managerial justification for the attributes selected should be provided). Each product attribute should consist of a minimum of 3 levels. A total 9-12 choice tasks should be included in the experiment and each task should include three concepts (options) and a “None” option, following the orthogonal, balanced and minimal overlap design requirements. Students will collect a small data sample (a minimum of 25 respondents) and randomly assign each respondent to choice tasks. Basic demographic information such as gender, age, education etc. should also be collected at the end of survey (after Part II) as individual specific variables to analyze segment-wise effects. Using the Sawtooth software, students need to analyze and interpret the results of attribute utilities, attribute importance, segment differences. Students are also required to simulate three concepts, report and interpret their choice probabilities with and without “None” option. Simulation insights such as segment effects, recommendations and sensitivity should be explored. The survey should be included in the report as an appendix.
Basic demographic information like gender, age, education, etc. should also be collected at the end of the survey (after Part II) as individual-specific variables to analyze segment-wise effects. Students need to look at the results of attribute utilities, attribute importance, and segment differences using the Sawtooth software and figure out what they mean. Students also have to simulate three ideas, report their results, and explain how likely their choices were with and without the “None” option. Simulations should be used to find out things like segment effects, recommendations, and sensitivity.
Part II
In the same Sawtooth software, after the CBC experiment (before the demographic questions), students then should use the same product, attributes and levels in the CBC to design a Conjoint Analysis study that measures the same respondents’ rating preference toward products profiles (create a set of Conjoint Analysis survey questions). Students should follow the below specific steps to design the study, collect and analyze the data.
1. Using the ‘Orthogonal Design’ in SPSS, generate a fractional factorial design that meets the orthogonality requirement. You can use the minimum (default) number of experimental/design product profiles generated by the SPSS, for example, it generates 9 product profiles with a 3 x 3 design. In addition, 2 holdout product profiles should be created.
2. Manually create survey questions in the Sawtooth for the experimental/design and holdout product profiles. You can measure rating preference by adding ‘Select’ or ‘Grid’ questions, each product profile should appear as a separate survey question.
3. After collecting the data, download the data in Sawtooth. The data file should be in the Excel format (.xlsx). Import the Excel data to SPSS by selecting File > Import Data > Excel…
4. Keep the rating-based conjoint data and remove the rest in the SPSS data, run a conjoint analysis by using the SPSS syntax learned in Week 9. Interpret the results, compare attribute utilities and attribute importance with those obtained in the CBC part.
5. Run simulations for exactly the same three product concepts simulated in the CBD study, compare the choice probabilities for the three product concepts.
For the above two parts, based on the summary of key findings, students should develop relevant and valid managerial recommendations for marketers with clear justifications based on findings. Assumptions and limitations of method should be appropriately addressed and clearly responded to the method.
Please attach the survey as an appendix and upload the Excel data file of the survey to iLearn submission link.
Assessment Task 2 – Marking Rubric
Criteria Fail 0% – 49% Pass 50% – 64% Credit 65% – 74% Distinction 75% – 84% High Distinction 85% – 100%
Articulation – Introduction and Background (10%) No or very poor introduction and description of the theoretical background of the proposed method Brief introduction and vague explanation on the theoretical background of the proposed method Brief introduction and limited explanation of the theoretical background of the proposed method Clear introduction of the proposed method and sound review of the literature to discuss the theoretical
background of the proposed method Succinct, precise and clear introduction of the proposed method, complete and insightful literature to present the theoretical background of the proposed method
Position – Research Design (30%) No or very poor design of the conjoint experiments Ineffective design of the conjoint experiments with some errors in attributes and attribute levels.
Product profiles are not completely or properly presented Effective design of the conjoint experiments, attributes, attribute levels, and product profiles are not fully presented. Effective and correct design of the conjoint experiments, attributes, attribute levels, and product profiles are fully presented Effective and precise design of the conjoint experiments, attributes, attribute levels, and product profiles are fully presented with theoretical justification
Analysis – Method (25%) Analysis uses none or
incorrect method. Analysis has many
errors. Correct analysis of the
majority of method. Correct analysis of the
method. Correct and complete analysis
of the method.
Interpretation – The key findings from the analysis (25%) Many errors/incompleteness in interpretation of statistical results Summary of results. Weak or incorrect interpretation of statistical results Summary of results. Detailed interpretation of statistical results with some errors. Summary of results. Detailed and correct interpretation of statistical results Summary of results. Detailed and correct interpretation of statistical results. Provided reasoning (what could be the reason
behind the observed results)
Critique – Conclusion and recommendations (5%) None, or confused and inconsistent with the key findings. None, or
irrelevant A brief summary of the key findings, with some relevant A brief summary of the key findings and their implications.
Relevant A clear summary of the key findings and their implications.
Relevant A succinct, precise and clear conclusion summarizing the key findings and their
implications. Relevant and
recommendations for marketers recommendations for marketers recommendations for marketers with some justification by relating it to findings (i.e. why
is it recommended) recommendations for marketers with justification by relating it to findings valid recommendations for marketers based on findings
Critique – Assumptions and None, assumptions Assumptions and Assumptions and Assumptions and Assumptions and limitations
limitations (5%) and limitations are limitations of method limitations of method limitations of method of method are appropriately
irrelevant to the identified have weak identified could affect are appropriately identified and clearly
method effects on the analysis the analysis results identified responded to the
results interpretation of the analysis
