Why is it critical to consider both the format of your data and the question(s) you wish to answer before selecting a statistical technique to use?

Policy: Why is it critical to consider both the format of your data and the question(s) you wish to answer before selecting a statistical technique to use?

The type of data you collect will determine what tests you can run on it. For instance, you can measure percentages, mode and chi-square for nominal data. You can measure percentile, median, rank-order correlation, or Friedman ANOVA for ordinal data. You can measure mean, standard deviation, product moment correlations, t-tests, ANOVA, regression, or factor analysis for interval data. And you can measure the mean and coefficient of variation for ratio data (Aaker, D., Kumar, V., Day, G., and Leone, R., 2010, p. 295).

 The question you are asking will determine what type of statistical technique you use. First state the problem. The steps for hypothesis testing are: 1) clearly state a null and alternate hypothesis, 2) choose the relevant test and a probability distribution, 3) choose the critical value or significance, 4) state the decision rule, 5) collect the data to test the hypothesis, and 6) make a decision and determine if the difference is statistically significant or just normal variation (Aaker, et al., 2010, p. 463). If you wish to determine if a population parameter is higher or lower than a specific value, you would use a one tailed test. If you wish to determine if a population parameter falls between two specified values, you would use a 2-tailed test. If sigma is known, or for very large samples, you would use a z-test. If sigma is not known, or for small samples, you would use a t-test. To determine the effect of an independent variable on a dependent variable, you would use a regression analysis (Aaker, et al., 2010, p. 523). œThe parameter 
b
1
 indicates that if the variable X is changed by one unit, the variable Y will change by 
b
1
 units. Thus, if $1 is added to the advertising budget, regardless of the level at which the budget is set, an extra 
b
1
  

customers will be expected to visit the store (Aaker, et al., (2010), p. 527).

Reference:

Aaker, D., Kumar, V., Day, G. and Leone, R. (2010). Marketing research. Hoboken, NJ: John Wiley and Sons

This is a response to the discussion question: 

Policy: Why is it critical to consider both the format of your data and the question(s) you wish to answer before selecting a statistical technique to use?

I need a page here (double spaced) 

Please make sure to write the response as if I was writing it. I’ve given you examples


 
¦

Click here to have a similar A+ quality paper done for you by one of our writers within the set deadline at a discounted.

Data format and question implications

Note # 1

by Kathleen Lee Myers

March 27, 2014 8:08 PM

_______________________________________________________________________________

Why is it critical to consider both the format of your data and the question(s) you wish to answer before selecting a statistical technique to use?

It is critical to consider both the format of your data and the question(s) you wish to answer before selecting a statistical technique to use because of how researchers what to report the results. For example, if researchers have used closed ended questions, a table of the number of each of the answers would be appropriate to report on (i.e. yes or no). However, if the questions that were asked were open ended, it may be more appropriate to report conclusion of the most popular answers and how they relate to the research question.

Suzie

This is a response to the discussion question: 

Policy: Why is it critical to consider both the format of your data and the question(s) you wish to answer before selecting a statistical technique to use?

I need a page here (double spaced) 

Please make sure to write the response as if I was writing it. I’ve given you
Conference 7

Note # 2

by Susan Smith

______________________________________________________________

Many statistical techniques may be used in both multivariate and univariate situations, but it is up to the researcher to determine which statistical technique to use when, based on the research hypothesis and the data collected.

Univariate techniques (techniques that measure a single or data independent of one another) can be used on both nonmetric and metric data. Remember, œnonmetric data are measured on a nominal or ordinal scale, whereas metric data are measured on an interval or ratio scale (Aaker, Kumar, Day & Leone, 2010, p. 453). Nonmetric data is assessed using nonparametric statistics, such as Chi-square and ANOVA, depending on if the researcher is using a one same or multiple independent samples. Parametric statistics, such as t- and z-tests, are used to examine the samples available for metric data.

On the other hand, multivariate techniques may be needed if œthere are two or more measurements of each observation and the variables are to be analyzed simultaneously ”such as a before and after measurement (Aaker, Kumar, Day & Leone, 2010, p. 453). Like univariate situations, there are multiple tests available in a multivariate setting: ANOVA, multiple regression, conjoint analysis, MANOVA, canonical correlation, factor analysis, and cluster analysis”to name a few.

Reference

Aaker, D.A., Kumar, V., Day, G.A. & Leone, R.P. (2010). Fundamentals of data analysis (pp. 439-459). Marketing Research (Tenth Edition). Hoboken, NJ: John Wiley & Sons, Inc.

This is a response to the discussion question: 

Policy: Why is it critical to consider both the format of your data and the question(s) you wish to answer before selecting a statistical technique to use?

I need a page here (double spaced) 

Please make sure to write the response as if I was writing it. I’ve given you examples

March 28, 2014 11:00 AM