Market Research Questions Question 1:

Market Research Questions Question 1: Give two specific examples for each type of the four types of primary scales. Explain how you would use each in a research project.

The four types of primary scales are the nominal, ordinal, interval and ratio. Nominal scales are used only for defining labels for variables, and have no statistical strength or value in the context of statistical analysis. They are primarily used for defining the location of a survey respondent or their sex, and are typically used exclusively for classification purposes. The next type of scale is called ordinal are often used for defining the relative distance between two or more attributes or entities, and these entities can be attitudes or beliefs in addition to behaviors. Two examples of ordinal scales are the use of the Likert Scale for measuring attitudes, and second, the measuring of preferences for one product or service over another. The third type of scale is an interval scale, and this specifically provides the ability to complete common statistical calculations including the mean, median, mode, and standard deviation of a series of values. As the name suggests, this scale quantifies the precise interval distance between one value and another. Interval scales are often used for defining exact magnitudes between data elements. This could include pricing variations by customer segment and the defining of market segments using advanced statistical techniques. The last and most statistically robust scale is the ratio scale, and this is unique in that it has a zero point. As the most statistically robust of scales, extensive analysis of data sets can be used for finding reliable correlations in data to predict pricing differences, behavioral responses to marketing campaigns, and the impact of pricing and promotion on sales. Question 2: Your new supervisor has asked you to evaluate the different types of scales and recommend one for purposes of a new product. Recommend one of the scale types to her, explaining your rationale for choosing it over the others.

I would immediately recommend the interval scale or higher due to the fact that these two scales (interval and ratio) have the ability to calculate distances between data points - or as is known in statistical calculations - they have an orthogonal property to them. Further, the predictive power of data based on the interval or ratio scale is superior, and also lends itself well to segmentation development using advanced statistical features including cluster analysis. From the perspective of measuring the new products' perception in a target market, I would also rely on interval scales as a minimum as this provides for defining both the optimal point of preference by potential consumers of the product and their ideal point. From these two data points, a perceptual map could be created that reflects how the new product is being perceived. I would tell my boss that at a minimum the need for having an interval-based scale was critical, and if possible and practical, ratio-based data especially on pricing and demand- based variables would be excellent to have. Question 3: Your supervisor has provided you with an instrument that was used in the past. She wants you to establish validity for the instrument. Of the ways you could assess this, pick a method and explain why you would use that method versus the others available to you.

My immediate question to the supervisor would be what specific purposes the questionnaires were used for to begin with, what were the objectives of the study, and what the results were. I would also want to know the sampling frame, the response rate, how the survey was administered, the resulting decisions made from the data, and the overall impact of the information gathered. I would also specifically want to understand how the scales used in the instrument (or survey) were specifically developed. I would then evaluate the results obtained versus the instrument used to see what the effectiveness of the instrument was, and finally apply a series of statistical tests to the results to see if they were accurate or not. Question 4: Your supervisor has provided you with an instrument that was used in the past. She wants you to establish reliability for the instrument. Of the ways you could assess this, pick a method and explain why you would use that method versus the others available to you. There are many approaches to ensuring the reliability for a survey instrument, with the most common being test-rest, split halves, and internal consistency. All of these approaches yield a scale between 0.00 and 1.00 with the higher the score, the higher the reliability. In a test- retest approach to measuring reliability is based on giving the survey to a group of respondents once, and then giving them the same survey a second time, then correlating the results to see the level of reliability of the instrument. Obviously the higher the correlation the greater the reliability, and despite the effects of learning on the context of this text, it is used very often for testing instrument reliability. The second approach, split-halves testing, is as the name suggests an approach of splitting a survey in half and then correlating the two halves. Splitting a survey and then measuring the resulting correlation is an effective technique for ensuring reliability as well. Finally there is the approach of internal consistency, and this approach focuses on the average correlation between all survey items and the number of survey items, to provide an estimate of reliability. Of al these methodic, I would choose the test-retest approach and explain that it is superior to the internal consistency approach, and that the split-halves approach can introduce too much variability into the process. Question 5: You have been tasked by your supervisor to design a questionnaire. What steps would you use in designing this questionnaire? Be specific, including exactly how each step of the process would apply to the construct you want to measure. In defining the steps for designing a questionnaire, the following are the series of steps I would use: 1. Clearly and precisely define the objectives of the survey first, being sure to specifically describe the results required resolving a problem or generating insights to guide a decision. 2. Define survey strategy as being either primarily qualitative or quantitative in focus. This makes a major difference on the structure of each question from this point forward. Qualitative is typically open-ended and in search of new information. Quantitative is more focused on measuring features, attributes, attitudes and preferences with greater precision. 3. Begin drafting questions that support the attainment of the study's objectives. Defining the scales for each question are also critical and require tying back to the specific objectives of the survey. 4. Align the survey questions so there is a logical flow from the standpoint of the respondent. 5. Test the survey with key respondents and validate results from this test through initial statistical analysis. 6. Create the survey and launch it as part of the broader research effort.

Question 6: For the questionnaire you were tasked to design in the previous question, discuss the type of items you would use and explain why you would use that particular type. Include the construct you are trying to measure, and how your item would help you to successfully meet that goal.

In the process defined in question 5, I would first focus on the need for quantify the perception of a products' features, functions and their association with the benefits that customers had received from them. To accomplish this I would use a Likert, 5 point scale to ensure that interval- level data was captured, so that perceptual mapping and ideal point distancing could be completed as a baseline for segmentation efforts. I would also want to see, at an interval scale level, the differences in perception between competitors' comparable products and my own. Finally, I would want to measure, again at an interval or ratio-level, what the ideal set of product features are for the customers surveyed and see how to design products specifically to accomplish that product through development. Question 7: You have administered a questionnaire, but only got a 10% response rate. Discuss how you would move this response rate to a more acceptable level. First, I would consider the respondent base that was given the questionnaire and critically analyze if they are the best possible respondent base of the survey itself. After evaluating this issue, I would next look at which responses I did get from the respondents and see what their demographics, role-based responsibilities are (especially in a B2B- based survey), and what their unmet needs were. Finally I would look to see which questions were and were not answered. All of this data would be used to evaluate how to better organize the research instrument and also select a more appropriate sample frame. If all these aspects were correct, I would include an incentive of a $5 Starbucks' coffee card in addition to offering the opportunity to win a personal DVD player. This would significantly increase the response rate, but I would be sure the underlying aspects of the survey were correct first.

Question 8: Define a hypothetical situation in which you would use an observation form in place of a questionnaire. Discuss the elements of this situation that would make it more practical for you to use the observation form.

In the case of researching the development of an entirely new class of automobiles, where people often don't know what they want until they see it, I would rely on observation research. These are typically called new car clinics, where potential new car models are previewed and car companies ask for feedback on how they are perceived by the public. If I was designing an entirely new SUV, I would go and observe how families use their SUVs today, and how singles use SUVs for their mountain camping hobbies. In short I would go and use observation research to see what unmet needs people had with SUVs and cars because asking them would yield little if any insights. For this type of research the interaction between people and products are critical.