Making Research Decisions
User Rating: 0 / 5
- Details
- Hits: 12235
Question one
- In developing an estimating equation that would be used to predict which applicant would come to our university as students, I would use frequency distribution as a method of multivariate technique (Alvarez & Olivera-Smith, 2013). I would choose on frequency distribution because it provides the applicant’s information in details. Using this method, the results would be found using analysis of variance test (Anderson & Shattuck, 2012).
- In predicting family income using such variables as education and stage in family life cycle, I would use multiple regression analysis technique as a multivariate method (Bandura, 2012). I would choose multiple regression analysis because it would help me examine the relationship between the income, a single metric dependent variable and education as well as stage in family cycle that are metric independent variables. The method would carefully observe the assumptions of normality, linearity and equal variance since it determines linear correlation taking into account the lowest sum of squared variances. This is a good forecasting tool (Bates & Sangra, 2011).
- If I wanted to estimate standard labor costs for manufacturing a new dress design, I would use conjoint analysis multivariate method, which is best fit for evaluating objects and relevant levels of the attributes requiring examination (Alvarez & Olivera-Smith, 2013). In this case, the object would be the new dress design and the various levels that the design must go through during the manufacturing process. This method helps to calculate utility for each level of manufacturing. Combination of attributes at every level is summed to obtain the overall preference for the attribute at each stage of manufacturing (Anderson, 2012).
- I have been studying a group of successful sales-people. I have given them psychological tests to find meaning out of these tests. I would use structural equation modeling multivariate method. This technique enables simultaneous examination of multiple relationships between sets of variables. The method is best for many scaled attributes (Bandura, 2012).