Statistics in Public Relations - Essay Example

and # Statistics in Public Relations Why is statistics important to Strategic Communications? Statistical data and analysis is important to Strategic Communications because this is a way that businesses can determine what type of decision to make regarding a product development strategy or the best ways to market to specific groups of consumers to get better sales. As one example, an educational publishing company, which is getting ready to bring out a new edition of a textbook, must determine first what consumers (teachers and students) liked, or did not like, about the current edition (Churchill and Iacobucci 565). The data gathering process is conducted through online questionnaires or surveys in order to determine the information, which is gathered and then analyzed. The Strategic Communication is first important in communicating to authors what was uncovered in the research. Secondly, when the book is ready to be sold, the Strategic Communication, through advertising and public relations (PR), reflects the changes made in this new edition to all those who potentially will adopt or purchase this book for their courses (Churchill and Iacobucci 566). As another example of Strategic Communications, a Vice-President in a business may need to inform the board of directors about a new process that can be implemented, but with some additional costs. The research done as to the plan’s effectiveness and what it will cost, is essential to selling this new process to the board of directors. Therefore, the statistics which shows this information, is an essential part to successfully selling the new plan or operation. 2. Differentiate between a population and sample. The difference is that a population consists, for example, of all people who drink coffee at Starbucks (Zikmund 296). The sample of the population must be taken from within that population of all people who drink coffee from Starbucks (Week 4, p.2). This may be one hundred people who have been qualified (or vetted) as being appropriate to take the online questionnaire or survey. Part of making sure that these people are representative of the target group is making sure in the application that they do buy Starbucks coffee at least once a month, for example. This is part of variables when we determine these parameters of the data while canvasing those people who have applied to take the survey Week 4, p.4). If any of the required parameters (variables) are not present, such as a certain age group, buying Starbucks coffee more than once a month, who are currently working, and own a home, for example, then those returns are not vetted or accepted. In general, use 10% of the overall group to get a representation, such as 100/1000 (Muijs 35). In describing a sample of School of Communication students, they would need to be taking classes in public relations, journalism, marketing, and radio/television courses (Zikmund 296). Age is not considered a factor but it can be a descriptive of the sampling, as older students also take these courses, not just those who are 18 to 22 years of age. In defining parameters for attaining the sample, the objective for this might be that only those who are majoring in any of these courses, and within a certain age group, can apply to the research study. Other parameters that could be applied in some cases, is to only have honor students apply, rather than the whole population of students in the School of Communications. Creating such parameters in the invitation communication means that only those who closely fit the parameters, will apply, rather than the whole school or college, thus saving on wasted time weeding out those who do not fit in to the objective (Muijs 35; week 4, p.6). 3. Explain what statistical significance is. Statistical significance is the process of testing data through various operations. One example might be in creating two different messages, with a call to action, which will be sent to a sample group. This would be called A/B testing to see which message produced the most results. A hypothesis may be designed to say that message A will produce more results in the call to action than message B, based on the offering provided for signing up or making a purchase, but you need to prove that through testing and then statistical analysis of results (Kopecky web; Zikmund 344). In statistical significance, the level of significance determines the choice between a null hypothesis and the alternative hypothesis, with a probability level of .05 or .01. As a consequence, the null hypothesis would be rejected if the probability of occurrence in the data observed, is lower or less than the significance level (Zikmund 345). It helps to have the hypothesis in place so that it can be verified, such as whether A message version performs better because the call to action offers a bonus. The B message does not contain a bonus and consequently, it is hypothesized that A message will outperform B message, based on the bonus offered. In returns that verify statistical significance, there will be a message output (SPSS) that states statistical significance, for example, is with 95% confidence (Kopecky web). Ultimately, statistical significance tests hypotheses for whether hypotheses set for the research study, are true or false (Week 5, p.6). This also provides the answers to questions within the study and provides the significance levels of impact of X on Y, for example. However, there can also be a case of X does not impact Y, but creates Z instead (Week 5, p.4). 4. Discuss hypotheses, research questions, regions of rejection, areas under the normal (statistical) curve, ?, and power in testing hypothesis and answering research questions. Give examples of each. Every research study conducted must have an objective, or a hypothesis, in order to first structure the questions properly to obtain the right data for subsequent analysis (Muijs 13). As an example, a hypothesis of Communications students may show that those who take the statistics class, are also honor students. Setting the alpha (?), for example, at .05, means there is an acceptable probability of 5% for Type I (Week 5, p.6; Zikmund 296). Therefore, regions of rejection in this case is that there are many Communications students who take statistics, but who are not honor students. That analysis will either be NULL which means there were no differences, or there were, indeed, differences found. In the case of testing two messages, A and B, the question to be answered is whether the bonus added (X) had an effect on people signing up or making a purchase (Y). If it did make a difference, then the research is two (2)-tailed, but if not, or there is another factor involved, then it is one (1)-tailed, which may be a problem in falsely accepting any differences in analysis. This is implied by a Type I or Type II error or probability of error. This are related to sample size and the preference is a minimal Type I error. Type II error indicates that there are no differences when, in fact, there are. When using the One or Two-tailed tests, the statistical curve will be different. A two-tailed test curve will show the normal curve balance of 0.25, for example, on either side of the curve at the same exact spot. Therefore, the middle is defined as exact and indicates a standardized normal distribution (Zikmund 304). The One-tailed test curve shows the result as skewed to one side of the curve only (Week 5, p. 8). The normal curve is the measure of central tendencies which describe the measures of dispersion. The power in testing hypothesis is that it determines whether there was enough in sampling to determine accuracy of the hypothesis (Week 5, p. 13). Part II: Scoring Data Description Screen shot I – Data View Screen shot 2 – Variable View (data has not been changed or coded) Written Data Analysis What the data shows is that there are 10 respondents being reviewed for scoring levels and the descriptive is by gender and by age. There are 6 females and 4 males. The age range is between 17 and 30 years of age. One female is 17 years old; one female and one male is 18 years old; one female and one male is 19 years old; one female is 20 years old; one male is 21 years old; one male is 25 years old; one female is 28 years old; and one female is 30 years old. The average age at 21.5 for the whole group. Separated by gender, average age of females is 22 while males is 20.75. In the Scoring 1 section, one female scored 98, with one male who came in at 90. The total average of all students was 81.11. The average for females was 75.6 and males was 88. The lower scores in the female section (with a larger population group of 6) pulled down the average in scoring in Score 1. One 18 year old female did not score at all which also accounted for the lower score average. In Score 2, one female scored 100 and one male score 95 as the top scoring individuals. Males, once again, averaged at 88.25 while females averaged at 78.6. The total group average was 82.88. One 20 year old female did not score at all which also accounted for the lower female score average. In both cases of scoring the one female in each Score session that did not score, assisted in dropping the score down below the male Score, even though it was a woman both times who got the highest score. However, for males, the scores were higher and all scored, therefore there was a higher scoring average among the males. Resources Churchill, Gilbert A. and Dawn Iacobucci. Marketing Research: Methodological Foundations. Fort Worth: Harcourt College Publishers. 2002. Print. Kopecky, Juliette. “A Marketer’s Guide to Understanding Statistical Significance.” HubSpot Online (Blog). 12 April, 2013. < http://blog.hubspot.com/marketers-guide-understanding-statistical-significance>. 28 September 2013. Muijs, Daniel. Doing Quantitative Research in Education with SPSS. 2nd ed. Thousand Oaks: SAGE Publications. Zikmund, William G. Essentials of Marketing Research. Philadelphia: Dryden Press. 1999. Print. Week 1, “Introduction to Statistics and Research in Advertising & Public Relations.” PowerPoints Class Handouts. Public Relations Course. 2013. Week 2, “Measurement.” PowerPoints Class Handouts. Public Relations Course. 2013. Week 3, “Descriptive Statistics – Describing Distributions.” PowerPoints Class Handouts. Public Relations Course. 2013. Week 4, “Predicting Parameters – Population and Sample Distributions and Statistical Inference.” PowerPoints Class Handouts. Public Relations Course. 2013. Week 5, “Testing Hypotheses.” PowerPoints Class Handouts. Public Relations Course. 2013. Read More