ABSTRACT
The analysis of data is a key skill of the marketing manager. An ability to understand basic methods of data analysis is useful. Data analysis can be done easily now, using computer packages such as Excel and SPSS
Before data is processed, it is assessed for completeness and coherence. The editing process involves computer or manual checking of the data, to look for respondent or interview errors or inconsistencies
Coding is the process that allocates a number to each answer and it is this that allows analysis to take place. Coding open questions involves using a sample of the completed questionnaires and developing a coding frame, or a list of codes for all possible responses to an open question
Data entry may be carried out automatically through CAPI, CAWI and CATI systems, or scanned into the computer using optical character recognitions software, or they may be entered by hand. Once this is complete, the data can be analyzed
There are four types of data that can be analyzed:
These refer to values that are given to objects that, in themselves, have no intrinsic numerical value. For example, we assigned a value to gender: 1 for men and 2 for women
We can count them and create percentages
These data represent rank order data. They do not imply that there is an equal gap between items ranked and there is no other meaning to them other than rank order
It is rank order data in which the intervals between the data are equal. These are also known as interval scales. Interval scales rank elements relative to each other, but not from any observable origin. This means that the data has its meaning only by virtue of the comparison between elements selected
Ratio data has an absolute zero or observable origin. For example, shoe size, products bought, or age. This means all analyses are possible
n The data from a sample will always be subject to error. We cannot be sure that the difference between two results is a real change in those values, or simply a result of the sampling error. If the difference is large enough not to have occurred through chance or error, then the difference is defined as statistically significant
n Regression analysis is concerned with dependence. For example, sales volume may be predicted based on other variables. The allocation of dependent and independent variables is more important in regression analysis. Movement in the dependent variables depends upon movement in the independent variables
n Sales forecasters use regression analysis. However, it is clear that the movement in a market is caused by a number of factors and this is dealt with through multivariate techniques, which we will look at later
This is the most common approach to regression. Least squares identifies a line of best fit between observations and this enables an estimated regression function that indicates the relationship. Simple regression analysis may be enhanced through the coefficient of determination. This measures the strength of the relationship between variables
Factor analysis reduces a large number of variables to a more manageable smaller set of factors, based on the interrelationships between them. It provides insight for the groupings that emerge and allows for more efficient analysis of complex data
This technique groups objects or respondents into mutually exclusive and exhaustive groups
The technique is often used in data base marketing to create segments, based on behaviour across a range of variables
Consumers rate objects, often brands, by the relative strength of an attribute compared to other objects or brands. This creates a perception of a ‘position’ in the market and is very useful for determining brand perception and repositioning
Conjoint analysis is a way of looking at customers’ decisions as a trade off between multiple attributes in products or services. In conjoint analysis, consumers are asked to make decisions about various attributes, trading lower price for comfort, for example, in car purchases
There are many software packages on the market that will do most of this for you
The key thing is to understand what these packages will do to your valuable data and to produce efficient analysis, which allows a focus on the research problem. Excel is adequate for most of the key formulae outlined above, but there are specialists; perhaps the best known software packages include: SPSS www.spss.com and SNAP www.mercator.co.uk
n Market metrics are used in business planning and marketing monitoring to keep the marketing programme on track. The most common market metrics that companies use are:
Market size
Market share
Market penetration
Installed base
Product usage
Customer attitudes
Brand awareness
Advertising awareness
Brand image
Customer satisfaction
n Quantitative surveys mean getting people to answer fixed questions in questionnaires. Because the objective is measurement, it is important that all people answer the same question
Go to www.cimvirtualinstitute.com and www.marketingonline.co.uk for additional support and guidance