Previous literature Conduct exploratory research 9 Wording needs to be precise

• Previous literature • Conduct exploratory research

9 • Wording needs to be precise and self-explanatory • The norm is to have 5 to 7 values on your scale A couple of things to remember when using metric questions for your questionnaire For both categorical and metric questions, the answers are typically coded in numbers Sample Survey Questions • Gender: Male = 1; Female = 2 • Age: ‘Under 13’ = 1; ’13 – 17’ = 2; ’18 – 25’ = 3; ’26 – 34’ = 4; ’35 – 54’ = 5; ’55 – 64’ =6; ‘65 and over’ =7 • Likelihood to repeat buy: “Definitely will not” = 1; “Definitely will” = 5 Answer coding for sample questionnaire in the previous slide Customer ID Gender Age Likelihood to repeat buy 0001 1 5 3 0002 1 4 4 0003 2 2 2 0004 1 7 2 0005 1 3 5 0006 2 3 1 0007 1 2 2 What the collected data look like

10 This dichotomy of categorical vs. metric Categorical Questions Categorical Data Categorical Scale Metric Questions Metric Data Metric Scale Sample Survey Questions we saw earlier Categorical Question Categorical Question Metric Question Collected data that we saw earlier Customer ID Gender Age Likelihood to purchase 0001 1 5 3 0002 1 4 4 0003 2 2 2 0004 1 7 2 0005 1 3 5 0006 2 3 1 0007 1 2 2 Categorical data Categorical data Metric data Why should I care about this distinction between categorical and metric data?

11 • Metric – data that you can add, subtract, and compute averages for • Categorical – can NOT add, subtract, or compute averages for (even if it is coded as numbers, such as in marital status on the previous page) This is an important distinction because comparative research questions entails comparing these averages. Main difference between metric vs. categorical data (for this unit) 8/23/2017 32 Example to illustrate this distinction ‐ can you answer the following comparative research questions (RQ) with the data I just showed you a couple of slides ago? • In the population from which the data are collected, are men and women on average of the same age? • In the population from which the data are collected, do men and women on average exhibit the same likelihood to repeat purchase? The reason that you cannot answer the 1 st RQ is because age data in this case is categorical (age group), which means that you cannot compute its average So what is the solution if you want to ask the 1 st RQ? Either don’t ask this question, or collect age information using a metric scale (more on this later) Department of Marketing Scale & Scaling

12 What is a scale ? Scaling • Goal - to precisely quantify what is to be measured. • Definition - the process of developing a scale (i.e., by coming up with a set of descriptors to represent the range of possible responses to a survey question) • Examples – Rate the car on a scale of 1 to 10 – Male, Female (value ≠ numbers) 8/23/2017 35 Many open-ended questions also make use of a scale, it is just that the scale is implicit and the range of values are not explicitly stated Example • What is your annual income before taxes? ___ – Scale: AUD$0 to infinity • What mark do you expect to achieve for this unit (MKF2121)? – Scale: 0 - 100 • What is the temperature today? – Scale: Celsius or Fahrenheit 36

13 There are 4 types of primary scales 37 Primary Scales Nominal Ordinal Interval Ratio Categorical Metric • Nominal scale – no rank-ordering between values – Example: Which one of the following media influences your purchasing decisions the most?