Finally, ‘multiple-source secondary data’ can be used as a combination of documentary and survey secondary data. The most frequently used by students and researchers are secondary data from (i) well-established institutions.
Bank, Eurostat, OECD, NBER), or (ii) company information located in databases such as Amadeus which provides financial data for more than 18 million public companies in Europe.
There two important things to be considered when undertaking research using secondary data:
- How easily you can get access to the data. It is very common in the early stagesof research to be unaware of data availability which means that you need tolocate the type of data required and how to access the appropriate source.
- How you can evaluate the validity and reliability of the data, as you might collectdata that will not provide you with the right information for the research questions you seek to answer.
Interviewer-completed questionnaires are completed by the interviewer (researcher in our case) and are very common when you are conducting structured face-to-faceinterviews or telephone survey questionnaires. These practices are widely used inmarketing research. For example, you may wish to measure consumer perceptions towards a new ecological product and you plan to interview customers in large supermarkets using a structured questionnaire.
Collecting data for statistical analysis
Data for statistical analysis can be used to produce a range of statistics varying fromtables, charts and simple descriptive statistics to inferential, more sophisticatedstatistics such as group comparisons, correlations, linear regressions or even morecomplex statistical models (Saunders, Lewis & Thornhill, 2012, p. 473). Whatever the choice made for the collection of data (primary or secondary), there are somecentralissues that should be carefully considered when aiming at statistical analysis of the data collected. The most important are introduced below.
Descriptive statistics’ are numbers that describe specific properties of a sample of numerical observations/information/data collected using a primary or secondary data collection process. The purpose of conducting and presenting descriptive statistics is to reveal the fundamental characteristics of our sample of data. In comparison, employing inferential statistical analysis aims at the generalisation of findings: You can draw conclusions about the properties of the (unobservable) population our sample comes from. (This will be explored further in Week
Measures of central tendency
- Mean: the arithmetic average of a set of data in a particular sample (can be usedonly with ratio and interval variables).
- Median: the middle value of a set of data in a particular sample, assuming that values are arranged in size order.
- Mode: the most frequently occurring value in a particular sample. It can be usedas a measure of central tendency for all types of variables. Note that it is thevalue where the histogram reaches its peak.
Measure of dispersion
- Range: is a simple measure of dispersion and is defined as the difference between the maximum value and the minimum value in a frequency distribution.
- Variance: is defined as the mean of the squared errors (from the mean). Thestandard deviation (square root of variance) is mainly used as a measure of dispersion, however this applies to ratio or interval variables only.