Major Issues on Descriptive Statistics - Essay Example

?Chapter Four Analysis and Results Introduction The analysis of statistical data serves as an important part of the overall research process. It is possible for a researcher to collect valid data that may be misinterpreted due to errors in the analytical techniques being used. This risk assumes greater importance in research where the sample sizes are large such as in the current research problem. In order to ensure that the data being presented for analysis is competent enough, the researcher employs a number of techniques. One of these techniques is better known as internal consistency / reliability or internal validity of the statistical data. The contention behind internal consistency checks is to ensure that the research questions produce consistent responses in the case of the respondents. It is common practice to insert similar questions in research surveys in order to gauge whether respondents answer these questions similarly (Berg, 2001). Research in marketing often entails working with a number of different variables in order to establish the relationship between such variables. Typically the data collected in order to carry out social sciences or marketing research relies in large part on quantitative research backed up by qualitative research to fill out the missing gaps. The use of large sets of quantitative data such as in the current research poses a number of problems in itself. For one thing, there are a number of variables who could be related to each other and may have an impact on the overall hypothesis. Such relationships between variables may pose strong or weak connections in one or more variables that need to be investigated. A preferred method to carry out such investigations is to use a factor analysis (Creswell, 2009) to determine the degree of correlation between various variables. Often a number of variables may be related to each other such that other variables may also exert an influence on the overall relationship. This can only be determined by utilising a proper factor analysis which may be related to but differentiated from principal component analysis (Bartholomew et al., 2008) The current research has focused on eliminating a number of problems from data collection and analysis by relying on factor analysis and internal reliability examinations. Descriptive Statistics Descriptive statistics are utilised in order to describe the major characteristics of a data set (Dodge, 2003). The contention in utilising descriptive statistics is to summarise the data set for analysis. In addition, descriptive statistics ensure that the respondents for a study all fall into the sub group being studied. For example, in the current research the aim is to describe the brand image habits and perceptions of ordinary consumers from the middle class. In this case, descriptive statistics ought to ensure that the respondents come from the middle class in large part or else the responses may be markedly different since the brand perceptions may be markedly different between consumer groups. The major descriptive statistics related to the respondents for the current research are presented in the tables provided below along with explanations. Favorite Brand: Frequency Percent Valid Percent Cumulative Percent Valid APPLE 64 40.3 40.3 40.3 BLACKBERRY 19 11.9 11.9 52.2 HTC 9 5.7 5.7 57.9 LG 2 1.3 1.3 59.1 MOTOROLA 5 3.1 3.1 62.3 NOKIA 20 12.6 12.6 74.8 SAMSUNG 33 20.8 20.8 95.6 SONY ERICSSON 7 4.4 4.4 100.0 Total 159 100.0 100.0 The first issue of importance was to establish which brands consumers subscribed to in connection to the cellular phone market. A number of different brands are available that subscribe to the tastes of different market segments so it was important to establish how the current cellular phone market was distributed. The results from the survey indicate clearly that the largest market leader is Apple (40.3%) followed by Samsung (12.6%) although it must be recognised that the proportion of Apple users and Samsung users differs significantly. Nokia (12.6%) and BlackBerry (11.9%) can be seen in intense competition with each other following the market leaders for cellular phones. Other brands such as HTC, Sony Ericsson, Motorola and LG form around 15% of the overall market. These results clearly indicate that Apple is the strongest market leader compared to all the other brands of cellular phones. Respondents' Gender Frequency Percent Valid Percent Cumulative Percent Valid Male 77 48.4 48.4 48.4 Female 82 51.6 51.6 100.0 Total 159 100.0 100.0 The gender of the respondents for the study represents an important sample set characteristic since the responses from male and female shoppers may be different. Typically, male shoppers are more brand loyal since they tend to spend less time at shopping compared to women. In order to ensure that the current study presented a balanced overall approach in terms of gender, the respondents were chosen so that each gender represented nearly one-half of the overall sample set. As seen above, males represented 48.4% of the overall respondents while females represented the remaining 51.6%. It is common for research to present biased results in case that gender is introduced as a bias creating factor in the research (Hanson et al., 2005). The use of a balanced respondent sample size ensures that the research results are not affected by the introduction of a gender bias. Respondents' Age Frequency Percent Valid Percent Cumulative Percent Valid 18-25 55 34.6 34.8 34.8 26-35 57 35.8 36.1 70.9 36-45 35 22.0 22.2 93.0 46-55 6 3.8 3.8 96.8 56-65 3 1.9 1.9 98.7 66 or more 2 1.3 1.3 100.0 Total 158 99.4 100.0 Missing System 1 .6 Total 159 100.0 The current research has chosen to focus on cellular phone manufacturing companies. It would be typical to assume that the prevalence of cellular phone use would be higher in younger people especially young adults than in other age groups above or below them. The descriptive statistics presented above provide a valid picture of how cellular phone usage is distributed in Great Britain. The table above shows that the largest group of cellular phone users originates from the age bracket of 26 years to 35 years of age who represent 35.8% of the overall sample. This age group is followed closely by respondents aged between 18 years and 25 years who make up 34.6% of the overall sample size. People aged between 36 years and 45 years represent the third largest group as they represent 22% of the overall sample size. People aged above 45 years represent under 10% of the overall sample size which indicates the low prevalence of cellular phone use in older people. Moreover, another notable observation from the table above is the absence of cellular phone users under the age of 18 years. Great Britain has cellular phone users under 18 years of age but the current survey has failed to pick up this clearly since no respondent was under 18 years of age. It could be speculated that this could introduce some form of bias to the current research but the number of cellular phone users under the age of 18 years is insignificant enough to be ignored for the purposes of this study. Respondents' annual income Frequency Percent Valid Percent Cumulative Percent Valid 0 6 3.8 3.8 3.8 ?10,000 or less 54 34.0 34.6 38.5 ?10,001 -?20,000 50 31.4 32.1 70.5 ?20,001-?30,000 24 15.1 15.4 85.9 ?30,001-?40,000 13 8.2 8.3 94.2 ?40,001-?50,000 7 4.4 4.5 98.7 More than ?50,000 2 1.3 1.3 100.0 Total 156 98.1 100.0 Missing System 3 1.9 Total 159 100.0 Income plays a major role in describing the respondents since income differentiation may exhibit large differences in brand perception, image and brand loyalty. The respondents were asked to list their incomes by selecting a suitable range that their income fell into. The results are shown in the table shown above. As mentioned before, most of the respondents for this study belonged to the middle class since the amount of respondents earning more than ?50,000 were a mere 1.3% which cannot affect the overall responses significantly. Moreover, most respondents for this research fell into the income group of ?10,000 or less (34%) followed by respondents earning between ?10,000 and ?20,000 (31.4%) followed by respondents earning between ?20,000 and ?30,000 (15.1%). The majority of respondents fall into earnings groups that are closely related. Given Great Britain’s social support structure, it could be expected that most of these respondents possessed nearly the same levels of disposable incomes. This implies that most respondents would tend to shop the same kinds of brands (that were targeted for the middle class) since their disposable incomes would not afford them more expensive cellular phone brands such as Vertu. This ensures that the current research will be consistent in its results since the majority of respondents possess similar disposable incomes, thus similar spending characteristics, and branding effects on their spending habits and personal lives. Highest level of completed education Frequency Percent Valid Percent Cumulative Percent Valid Less than a high school graduate 4 2.5 2.6 2.6 High School Graduate 35 22.0 22.6 25.2 Undergraduate Degree 65 40.9 41.9 67.1 Postgraduate Degree 51 32.1 32.9 100.0 Total 155 97.5 100.0 Missing System 4 2.5 Total 159 100.0 Consumer research is highly influenced by the levels of education exhibited by the respondents. This is all the more true for products that require education to play a role in defining their usability (Leedy & Ormrod, 2010). Products such as cellular phones or laptops could only be expected to be used by people who can interpret how to use such devices. The table above depicts the level of education of the respondents for the current study. It can be seen clearly that the majority of respondents held an undergraduate degree (40.9%) followed by a postgraduate degree (32.1%) while a smaller proportion had only high school education (22%). The number of people holding less than a high school degree represent only 2.5% of the overall sample size which can be ignored for most purposes. Moreover, the distribution of education levels of the current respondents is out of line with the national averages which would display a higher proportion of people with high school education compared with people holding undergraduate or postgraduate degrees. This could be attributed to the location where the survey was carried out since the City of Hull is home to the University of Hull which means that more people are more highly educated when compared to the country as a whole. Internal Reliability Data collected for research purposes especially data collected for quantitative research may be inconsistent and could lead to misleading findings. In order to interpret the collected data as close to the actual situation, it is essential to ensure the consistency or reliability of the collected data. A number of methods and tests are used such as the split halves test, the Cronbach’s alpha test, the Kuder Richardson test etc. (Prior, 2003). The choice of which method to use and which test to apply is determined by considerations such as the computational resources required, the amount of time required for analysis, the expected amount of inconsistency and others (Rihoux, 2006). For the current research, internal reliability was ensured by supplying the respondents with similar questions throughout the survey. This ensures internal reliability if respondents supply similar responses to the slightly differently worded questions that seek to measure the same thing. In case that the respondents supply differentiated answers for similar questions, it is highly likely that the collected data is not internally consistent or reliable. For example, to question the importance of brands in the respondents’ lives the following two questions were used: 1. Part of me is defined by important brands in my life. 2. I can identify with important brands in my life. In order to test for internal consistency from the collected data, the split halves test was used. The survey was designed with two questions of a similar sort placed at regular intervals. The responses to both questions needed to be similar in order to pass the split halves test (Tashakkori & Teddlie, 2003). The results of the split halves test showed that the collected data was reliable since responses to similar questions were similar for more than 90% of the sample set. The high degree of consistency meant that more complicated internal reliability tests such as Cronbach’s alpha were not required (Teddlie & Tashakkori, 2008). Factor Analysis Factor analysis qualifies the degree of correlation between various variables used in research. Typically, marketing research entails a large number of variables so there is a problem in dealing with how these variables are related (Williams, 2007). Factor analysis was utilised in the current research in order to clarify the relationships between the various variables being studied. The contention was to establish how CBI was influenced by other factors especially in brand loyalty and brand promotion. Results obtained from the factor analysis are presented in the tables shown below along with relevant discussion. Brand Engagement in Self Concept I have a special bond with the brands that I like. Frequency Percent Valid Percent Cumulative Percent Valid Strongly Disagree 3 1.9 1.9 1.9 Disagree 10 6.3 6.3 8.2 Slightly Disagree 13 8.2 8.2 16.4 Neither Agree Nor Disagree 37 23.3 23.3 39.6 Slightly Agree 47 29.6 29.6 69.2 Agree 37 23.3 23.3 92.5 Strongly Agree 12 7.5 7.5 100.0 Total 159 100.0 100.0 The table above clearly indicates that around 30% of all respondents slightly agree about having a special connection with the mobile brand that they prefer. Respondents who neither agree nor disagree with the posed question (23.3%) follow those who slightly agree. Similarly, respondents who strongly agree with this statement constitute some 23.3% of all respondents. Put together, respondents who feel a special bond with their favourite brand are far more than respondents who disagree with this statement in any degree. This clearly indicates that consumers have a perception of a special bond with their favourite brand that might play a role in the brand’s enhancement. Brand Engagement in Self Concept My favourite brands are an important indication of who I am. Frequency Percent Valid Percent Cumulative Percent Valid Strongly Disagree 11 6.9 6.9 6.9 Disagree 25 15.7 15.7 22.6 Slightly Disagree 8 5.0 5.0 27.7 Neither Agree Nor Disagree 43 27.0 27.0 54.7 Slightly Agree 41 25.8 25.8 80.5 Agree 20 12.6 12.6 93.1 Strongly Agree 11 6.9 6.9 100.0 Total 159 100.0 100.0 The table indicates that about 27% of the respondents have a neutral feeling about their favourite brand while 25.8% of all respondents slightly agree with the posed question. Respondents who agree with the statement (12.6%) are less than those who disagree with the statement (15.7%) while those who strongly agree and strongly disagree with this statement are balanced out at 6.9%. This indicates a nearly bell shaped distribution of respondent’s responses for the posed question which means that brands as a means of personal indication have a weak connection to brand enhancement at best. User Imagery Self- Congruency The image of the typical user of [Brand X] is consistent with how I see myself. Frequency Percent Valid Percent Cumulative Percent Valid 0 1 .6 .6 .6 Strongly Disagree 6 3.8 3.8 4.4 Disagree 6 3.8 3.8 8.2 Slightly Disagree 13 8.2 8.2 16.4 Neither Agree Nor Disagree 62 39.0 39.0 55.3 Slightly Agree 45 28.3 28.3 83.6 Agree 21 13.2 13.2 96.9 Strongly Agree 5 3.1 3.1 100.0 Total 159 100.0 100.0 In respect to this question, we ask the sample about the image of the typical user of the brand they prefer and If it similar to how they see themselves. Around 39% of the shoppers were neutral and about 28% slightly agree with the posed question. On the other hand, 13.2% of all respondents agreed with the statement. Respondents who disagreed with the current statement are far lower than those who agree with the posed question. This indicates that consumers develop a certain image of how they see other consumers of the same brand making them agents of brand image enhancement. Moreover, consumers tend to create images of how they perceive the brand and anyone else who fits into a similar category is another user of the same brand for them. Consumer Brand Identification Bergami and Bagozzi Please indicate to what degree your self-image overlaps with [Brand X] image by circling the most appropriate number Frequency Percent Valid Percent Cumulative Percent Valid 0 10 6.3 6.3 6.3 Not at All 18 11.3 11.4 17.7 To a very slight extent 19 11.9 12.0 29.7 To a small extent 29 18.2 18.4 48.1 To a moderate extent 40 25.2 25.3 73.4 To a considerable extent 28 17.6 17.7 91.1 To a great extent 12 7.5 7.6 98.7 To an extreme extent 2 1.3 1.3 100.0 Total 158 99.4 100.0 Missing System 1 .6 Total 159 100.0 As per the Bergani and Bagozzi method, the table above indicates that most respondents agreed with the brand providing a self image overlap (25.2%). Responses to either side i.e. not providing such a self image (18.2%) or providing such a self image (17.6%) were similar. It can be seen from these results that consumers see brand image as providing them with a self image. My close friends or relatives warned me not to buy a competitor brand (to brand X) mobile phone Negative Word of Mouth Received Frequency Percent Valid Percent Cumulative Percent Valid Never 33 20.8 20.8 20.8 2 17 10.7 10.7 31.4 3 23 14.5 14.5 45.9 4 33 20.8 20.8 66.7 5 31 19.5 19.5 86.2 6 13 8.2 8.2 94.3 All the time 9 5.7 5.7 100.0 Total 159 100.0 100.0 The table above shows decisively that consumers of a certain brand are arranged socially in groups that follow similar brands. Around 20.8% of all respondents were never warned by close social contacts to buy their favourite brand which indicates the prevalence of similar brand image in social circles. I speak positively about [Brand X] to others Brand Promotion Frequency Percent Valid Percent Cumulative Percent Valid Strongly Disagree 5 3.1 3.1 3.1 Disagree 10 6.3 6.3 9.4 Slightly Disagree 9 5.7 5.7 15.1 Neither Agree Nor Disagree 38 23.9 23.9 39.0 Slightly Agree 45 28.3 28.3 67.3 Agree 39 24.5 24.5 91.8 Strongly Agree 13 8.2 8.2 100.0 Total 159 100.0 100.0 The table above clearly indicates that consumers who have any amount of inclination for their favourite brand tend to advertise it to others. However, only a few respondents spoke negatively of their brand which indicates that consumers are influenced by brand image enough to advertise the brand to others. Correlations Consumer Brand Identification Mael and Ashforth Value Congruency Consumer Brand Identification Mael and Ashforth Consumer Brand Identification Mael and Ashforth Pearson Correlation 1 .342** .535** Sig. (2-tailed) .000 .000 N 159 159 159 Value Congruency Pearson Correlation .342** 1 .427** Sig. (2-tailed) .000 .000 N 159 159 159 Consumer Brand Identification Mael and Ashforth Pearson Correlation .535** .427** 1 Sig. (2-tailed) .000 .000 N 159 159 159 **. Correlation is significant at the 0.01 level (2-tailed). Correlations Value Congruency Consumer Brand Identification Mael and Ashforth Brand Attractivness Value Congruency Pearson Correlation 1 .356** .285** Sig. (2-tailed) .000 .000 N 159 159 159 Consumer Brand Identification Mael and Ashforth Pearson Correlation .356** 1 .267** Sig. (2-tailed) .000 .001 N 159 159 159 Brand Attractivness Pearson Correlation .285** .267** 1 Sig. (2-tailed) .000 .001 N 159 159 159 **. Correlation is significant at the 0.01 level (2-tailed). Correlations Identity Salience Value Congruency Perceived Quality Identity Salience Pearson Correlation 1 .545** .360** Sig. (2-tailed) .000 .000 N 159 159 159 Value Congruency Pearson Correlation .545** 1 .318** Sig. (2-tailed) .000 .000 N 159 159 159 Perceived Quality Pearson Correlation .360** .318** 1 Sig. (2-tailed) .000 .000 N 159 159 159 **. Correlation is significant at the 0.01 level (2-tailed). The results of the correlations shown above clearly indicate that consumers of a strong brand tend to enhance the overall brand image by advertising it. The consumer brand image (CBI) influences the overall brand loyalty and promotion. It is clear that a large number of consumers see brand image as an extension of their personal self and feel insulted if their brand is being insulted. Moreover, consumers see any strides on the part of the brand as advances on their own personal part. This indicates the existence of brand loyalty on a very strong note in most consumers. In addition, brand loyalty is also confirmed by the relationship between personal quotes about the brand being quoted as “we”. The feelings of consumers that their favourite brand supports the same values as they do also indicates that a large brand loyalty exists that could drive brand promotion without a doubt. The attitude of consumers to indicate that their mobile phone brand is the best by a large part of the respondents indicates that brand loyalty is strong. Moreover, the idea that consuming a product is more than just buying a product shows that most consumers see their favourite brand as an extension of their personal self. CHAPTER FIVE CONCLUSION Discussion of Research Questions The results of the current research presented above clearly indicate that there is a strong connection between consumer self projection and brand image. Most consumers tend to see brand images as parts of their own self which indicates a deep intertwined relationship between both. When the research questions and the posed hypothesis are analysed in this regard, it becomes clear that brand image on the part of the consumer has a major role in brand loyalty and promotion. The first hypothesis speculates that consumers will play any role in enhancing the overall image of the brand and tend to promote the brand in this process. However, this hypothesis is not strongly supported as most consumers tend to linger between agreeing and disagreeing with the provided statements. On the other hand, the second hypothesis holds far more weight. The second hypothesis speculates that consumers who are able to identify themselves with the brand image are able to enhance the overall brand image more. This is truer compared to the first hypothesis since the results indicate that consumers who are more strongly attracted to a brand are willing to go to great lengths such as feeling personal insult or using “we” to refer to the brand in order to promote the brand. The third hypothesis is also supported in this regard since consumers who strongly identify with the brand tend to feel emotionally attached to the company. This implies that the second and third hypothesis are confirmed by this research while the first hypothesis is rejected. Recommendations The strong intertwined character of brand image and consumer loyalty is undeniable. The projection of this character is all the more strong for consumers with greater brand loyalty as part of their overall personality. The scale of brand loyalty that affects brand promotion is clear so companies need to look more into consumer loyalty than other means to get better reviews and hence more sales. The current trend in marketing to target the consumer with external agents such as media might not have the effect of personalised factors such as a person in the consumer’s social circle advocating the brand. Modern businesses need to concentrate on enhancing consumer brand image to support greater promotion. Conclusion The current research is indicative of the fact that modern marketing efforts require more support from already existing customers. The great divide between Apple and other cellular phone brands can only be attributed to more loyal consumers since Apple has more to offer than other brands. The projection of loyal consumers onto other people in their social setup means that a company or any brand has far greater chance of raking into another market segment through existing consumers than otherwise. The entry of Apple into the cellular phone market is recent compared to previous giants such as Nokia or Samsung but its market share is indicative of its marketing strategy. The future clearly shows that marketing will become more of a personalised, loyalist consumer effort than one driven by external marketing means alone. References Bartholomew, D.J., Steele, F., Galbraith, J. & Moustaki, I., 2008. Analysis of Multivariate Social Science Data. 2nd ed. New York: Chapman & Hall. Berg, B.L., 2001. Qualitative Research Methods for the Social Sciences. Needham Heights, Masachusstes: Allyn & Bacon. Creswell, J.W., 2009. Research design: Qualitative, quantitative, and mixed methods approaches (3rd Edition). Thousand Oaks: Sage. Dodge, Y., 2003. The Oxford Dictionary of Statistical Terms. New York: Oxford University Press. Hanson, W.E. et al., 2005. Mixed methods research designs in counseling psychology. Journal of Counseling Psychology 52(2), pp.224-35. Leedy, P.D. & Ormrod, J.E., 2010. Practical Research: Planning and Design, Ninth Edition. New York: Merrill. Prior, L., 2003. Using Documents in Social Research. London: Sage. Rihoux, 2006. Qualitative Comparative Analysis (QCA) and Related Systematic Comparative Methods: Recent Advances and Remaining Challenges for Social Science Research. International Sociology, 21(5), pp.679-706. Tashakkori, A. & Teddlie, C., 2003. Handbook of mixed methods in social & behavioral research. Thousand Oaks: Sage. Teddlie, C.B. & Tashakkori, A., 2008. Foundations of Mixed Methods Research: Integrating Quantitative and Qualitative Approaches in the Social and Behavioral Sciences. New York: Sage. Williams, C., 2007. Research Methods. Journal of Business and Economics Research 5(3), pp.65-72. Read More