Research Techniques - Case Study Example

Introduction Surveys are part of research programs, that are always premeditated as formal statistical supposition about some population of a specific data collected from a division of that population (Bachman, Elfink and Vazzana, 1996). There are three options in survey design, and these are contact mode, response mode and sampling methodology, these designs should be made and those options should be evaluated in light of the cost propositions and the succeeding impact of those options may have data quality and the accuracy of the survey limit calculates approximately. This research options must be clarified in the early stage of the research planning. In this period of globalization, innovations has introduced the Internet that have generated new methods in conducting surveys, these surveys are done through electronic mail or e-mail and the World Wide Web. In e-mail surveys, the content of the survey is in the original message of the e-mail, in this case the respondents can complete the survey just by answering the original e-mail (Bradley, 1999). While Web surveys, are powered by host websites, in this case the respondents can just click on the website then complete the survey. Internet surveys are presently in trend mainly because of four accepted statements about how they pile up against more-conventional survey methods (Coomber, 1997): (1) the responses are quick; (2) they are equally even better than conventional surveys (3) they are much economical to perform; and (4) they are easier to implement. On the other hand, these statement may be true or not, depending on the individual conditions of the survey. Internet Surveys In the most common sense, conducting market survey research using the Internet engages performance one or more events and examining the outcomes. Researchers perform testing for a broad range of basis, series from pre-testing a research design to challenging to process a measuring instrument. A widespread application engages examining fundamental hypotheses. Once suitable, a prescribed testing gives the most reliable proof of causation possible (Deacon ET. Al.:1999). To perform one, researchers must employ two things. First, researchers must influence at least one independent variable. One does not just watch and calculate an independent variable as it obviously takes place. As an alternative, it is restricted and diverse as part of the design of the study. Second, opportunity practices must decide whether or to what extent a person investigational contributor is depicted to the independent variable. This is also called as random assignment. The submission of numerous independent variables in investigational designs is significantly helpful. If the outcomes of understanding to media be at variance for diverse people, such creates consent to the knowledgeable study of these divergences (Jensen: 2002). For example, including both experiences to mediated antagonism and gender in one design permits a researcher to observe both the detach relation of each independent variable with hostility and their interaction. An interaction happens every time the consequence of one thing on another differs according to some third variable. For example, disclosure may have more impact on males than on females. For the reason that of their value for causal inference, investigational studies are high in internal validity, nevertheless, this strength carries a price. By description, they involve arranged backgrounds that may vary radically from the way viewers make use of mass communication in the "real" world. For an instance, contact to broadcasted violence usually happens willingly and at home (Malhorta and Birks: 2000). If a researcher be in command of it, one has no guarantee that the outcomes studied will be relevant in more distinctive contexts. Consequently, experimental research may specify what can happen, rather than what does take place in "realistic" situations. Experimental research also leans to be restricted to the short-term effect of the mass media. One typically cannot influence and have power over media-related activities for months or years. Possibly, the Internet has an imperative impact not only on people, but on social groups varies in size from families to the whole populations and cultures. Therefore, not like more conventional scholastic disciplines like psychology, Internet Market survey Research utilizes a wide assortment of levels of analysis. In research, a level of analysis refers to whatever a researcher studies the characteristics of (Newman: 1998). In the case of psychologists in general are apprehended with the behavior of a person, while sociologists characteristically study the characteristics of groups. A media researcher researching the impact of television violence on the aggressiveness of individual children utilizes a psychological level; an important person studying the impact of mass communication on the economic development of entire countries works at a sociological level. In general speaking, media research engages more on the psychological or sociological levels (Saunders, Lewis and Thornhill: 2003). The sociological level is tremendously broad, covering groups varying in size from dyads to complete countries. In fact, most media research is openly social psychological, a discipline usually described as the study of how individuals are pressured by other persons. Even within the discipline of social psychology, a quantity of researchers reveals a psychological turned and others a sociological one. In using the Internet in survey, research data analysis is guaranteed to be accurate because the answers can automatically update the companies’ Internet surveys give more in-depth survey results and analysis. At the same time, faster turnaround can be expected because online market survey research activates surveys and gives access to live data. Companies can be prepared to inform, modify or correct their marketing activities if urgently needed (Crawford et.al, 2001). Since Internet survey research has reporting capabilities, as an alternative of generating hundreds or thousands of outcome reports by hand, they can all be computerized by giving a single policy for report screening and analysis furthering reliability across global organizations (Coomber,1997). Problems Methodologies are perspectives on research; they set out a vision for what research is and how it should be conducted. They are the connection between axioms and methods; methods are tools-techniques of data gathering, techniques of analysis, and techniques of writing. Because it is a tool, a particular method can often be used by many different methodologies. Therefore, methodologies are at a more abstract (or general) level than are methods. Methodology is like a strategy-or plan-for achieving some goal; methods are the tactics that can be used to service the goals of the methodology. In essence, methodologies provide the blueprints that prescribe how the tools should be used. Those prescriptions can be traced to the axioms beliefs about how research should be conducted. Although, Internet surveys have advantages and very popular in conducting social research, Internet survey has numerous problems that needs to be taken care of and these are (Couper, 2000): The different levels of computer knowledge of the respondents may result to errors or non-response. The researcher may face challenges on data security on the website or server. There is no method for choosing randomly for samples in the Internet. The respondents may not be able to complete the survey, sometimes the system of the website be down or computers crashes. Internet surveys cannot be use to generalize population because not all people have the access to it. In order to solve the problem regarding security and privacy, the researcher may use user log files as an advantage for both researcher and respondent. The researcher may also give a description of the questionnaire design needed to gather his/ her data. Therefore, the issue on privacy will be a double-edged sword for both respondent and researcher. There are also website servers and providers that give membership email list that can facilitate sampling frame to the researchers. However, not all members of virtual groups and organizations allow their email addresses to be listed, and some may not allow administrators to provide their email addresses to researchers. This makes accurately sizing an online population difficult. The other solution to tracking of the replies of the respondents is through multiple responses system (Couper, 2000). Respondents are asked to send their email address in order to complete the survey; the program will remember the email address and does permit anyone using the email address to have an access to the survey. Therefore, multiple responses will be avoided. With an approximation of the quantity of bias, individual could then merge information from the ease and probability samples to give way to more accurate approximations than would be likely from the possibility sample alone. If the expediency sample is very biased, then it will be nearly useless. This shows that the probability segment of the sample would have to be large enough to stand on its own in a worst-case scenario (Bradley, 1999). Another source of problem in using Internet survey is the social bias If trying to get rid of the bias from the convenience sample will prove ineffective, then the only alternative is to use the (potentially) biased data in the estimation. However, as we show later in this appendix, and as one might expect, the bias of the convenience sample must be small. One way to take action to this inadequacy may be to focus on estimating parameters that are less subject to bias, such as within-subject differences or regression coefficients, rather than population estimates of proportions or means. One can also use post-stratification to reduce bias as much as possible. For example, a small set of questions can be integrated in both the convenience and probability samples that are (1) associated with likelihood of participation in the Web-based convenience sample (for example, age, education, computer use, and other such factors) and (2) likely to be associated with the parameters being measured. To use the post-stratification variables, one should treat the characteristics of the probability sample as the target and model the relative response probabilities of members of the “convenience sample pool” with given values of post-stratification variables. Weights inversely proportional to these estimated relative probabilities are then applied to the convenience sample only. The design effect from this procedure will decrease the efficient sample size (ESS) of the convenience sample, but the low cost of these observations makes compensating for moderate design effects on the convenience sample affordable References Babbie, E. R. (1998). The Practice of Social Research. 8th Edition. Belmont CA: Wadsworth Publishing. Bachman, E., J. Elfrink, and G. Vazzana (1996) “Tracking the Progress of EMail vs. Snail-Mail,” Marketing Research, Vol. 8, pp. 31–35. Bradley, N. (1999)“Sampling for Internet Surveys: An Examination of Respondent Selection for Internet Research,” Journal of the Market Research Society, Vol. 41, pp. 387–395. Coomber, R. (1997) “Using the Internet for Survey Research,” Sociological Research Online, Vol. 2, 1997, pp. 14–23. Couper, M. P.,(2001) “The Promises and Perils of Web Surveys,” presentation, RAND, Santa Monica, California. Crawford, S., M. P. Couper, and M. Lamias (2001) “Web Surveys: Perceptions of Burden,” Social Science Computer Review, Vol. 19, No. 2, pp. 146–162. Creswell, J.W. (1994). Research design: Qualitative and Quantitative approaches. Thousand Oaks, California: Sage. Deacon. D., Pickering, M., Golding, P. & Murdock, G. (eds) (1999) Researching Communications: A Practical Guide to Methods in Media and Cultural Analysis, London: Arnold Jensen, K. (2002). A Handbook of Media and Communication Research: Qualitative and Quantitative Methodologies. London: Routledge. Malhotra, N.K. & Birks, D.F. (2000). Marketing Research: an Applied Approach, 3rd edition, Essex, Prentice Hall. Marshall, C. & Rossman, G.B. (1999). Designing Qualitative Research. 3rd Edition. Thousand Oaks, CA: Sage. Maykut, P.S. & Morehouse, R. (1994). Beginning Qualitative Research: A Philosophic and Practical Guide. Washington, D.C.: Falmer Press. Newman, I. & Benz, C.R. (1998). Qualitative-quantitative research methodology: Exploring the interactive continuum. Carbondale, IL: Southern Illinois Saunders, M., Lewis, P. & Thornhill, A. (2003). Research Methods for Business Students. 3rd Edition. New York: Prentice Hall. Read More