Primary data is all about collecting and using data directly from the area of concern using methods such as interviews, surveys and focus groups while secondary data entails using data that is already existent like in journals and newsprints. Secondary data gives details about past areas of interest that were researched on while primary data gives information on the current area of interest to the researcher. Some of the disadvantages of using newspaper and magazine articles as secondary sources are that a person cannot know for sure if the articles represent the views of the writer or whether it represents a whole range of views. Secondly, the researcher intending to use secondary sources may find it difficult to ensure the data is reliable authentic or representative. Both primary and secondary data should be used in research as the secondary data gives the background information on the topic of research and sheds light on the relationships and conclusions of past research that blend into the current area being researched on using primary data in order to come up with concrete findings and recommendations.
Statistics presented on the evening news are valid and reliable but it all depends with the News Company or station as others are more reliable. It boils down to the origin, technique and methodology used to gather data. Most of the statistics are analyzed in real time like in opinion polls. However, there is always a marginal error in any statistical undertaking. Errors resulting from questionnaire design might affect the process of gathering relevant and accurate information resulting in a poorly conducted research at the end of it all. Research design and statistical information affect validity in that the research design shows how data is to be collected and which method is best for a particular topic of research. It acts as bedrock of data gathering. Statistical information is very sensitive as it involves figures and graphical representations of data. Its misrepresentation may result in huge errors and the invalidity of the findings of the research.
The advantages of using a survey is that they are relatively cheap and easy to conduct, they are reliable in describing the characteristics of a population. It is flexible in terms of the options to be used in a survey like face-to-face interviews, e-mail, telephone interviews and study groups. There is standardization in formulation questions and analysis of results. Another advantage is that it can cover a large area. The disadvantages of using survey are that it requires an initial study design that cannot be altered in the process of data collection. The researcher must ensure that a large number of respondents actually do reply. To do this the researcher has to come up with general questions that are appropriate for all respondents. Loaded questions are based on assumptions. An example of a loaded question is, “how often do you do that?” this assumes that the respondent does it sometimes. Focus groups is a type of quantitative research involving a group of people which is asked about their opinion, beliefs, suggestions, perceptions and attitude towards a research topic. Situations that are appropriate in using focus groups are: when a company wants to launch a new product, repackaging or rebranding of another product. The focus group is interviewed by a trained moderator in the group’s natural environment where they air their views and concerns. Another situation for using a focus group would be in social sciences where the researcher would be interested in the culture and beliefs of a certain community. Focus groups are beneficial in urban planning where the people are asked what amenities they require among other developmental issues.
Real world examples of a sample and a population would be a sample of students in a classroom within a high school, while a population would be inclusive of all the students in that high school. On a larger scale a sample would include a group of students in high school while the population would include all the students attending high school within a whole state or country. Non probability sampling does not give the samples an equal chance of being selected for being in the research hence it’s not random. Examples are convenience sampling where the samples are easily accessible to the researcher hence it’s the easiest form of non probability sampling. The other example would be judgmental sampling where the researcher sees certain people fit for the research than others. This is also known as purposive sampling.
Probability sampling gives all an equal chance of being included in the sample thus it is random in nature. An example is the simple random sample where all elements within a sampling frame have an equal opportunity of being selected. When sampling is done each element is selected independently. The other example would be by cluster sampling where the researcher selects groups of people who have similar characteristics and then selecting elements from within these clusters for the actual sample.
The best data scales suited for presentation of a pie chart is the ordinal scale. The pie charts that accompany a newspaper article are effective in explaining the statistics used in the articles because they give a visual representation and clearly outline differences in the presentations where they occur. A pareto chart would be used in companies for cost analysis, analyzing customer complaints, business processes and product defects.
All data does not have mean, median and mode because there are two types of data, quantitative data that has the above values and qualitative data that is concerned with non-numerical data. The mean is the best measure of central tendency when data is symmetrical and has no outliers but if there are, they should be included. The median is the best measure of central tendency when there are a huge number of outliers or when the data is skewed.
The relationship between research design and the operational definition of a problem statement is that the problem statement will determine the variables and the statistical methods to be used in the research design. Descriptive statistics are used to describe the features of the data in a study. They provide simple summaries on quantitative data by the use of either graphical summaries such as dispersion graphs and histograms or numerical summaries that are the measures of central tendency. The relationship between operational definition statement and descriptive statistics is that the former is a statement that describes the specific methods for measuring a variable while the former is a summary of data that has already been measured.
The introduction of Michael Jordan’s salary would skew the distribution because it consisted of a huge figure which is an outlier as compared to the other figures. The median would be the best measure of central tendency as it would show the middle value without being affected by the outliers and skewed data.