Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Standard view
Full view
of .
Look up keyword
Like this
0 of .
Results for:
No results containing your search query
P. 1
MArketing research notes chapter17

MArketing research notes chapter17

Ratings: (0)|Views: 100|Likes:
Published by manojpatel51
this are the notes for marketing research - a subject for TYBMS, mumbai university.
this are the notes for marketing research - a subject for TYBMS, mumbai university.

More info:

Published by: manojpatel51 on Mar 14, 2009
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





© Copy Right: Rai University
100 11.556
In this chapter we shall learn to apply techniques of statisticalinference to business situations. I will first introduce the basic ideaunderlying statistical inference and hypothesis testing. Our focuswill be more on understanding key concepts intuitively and lesson formulas and calculations, which can be done easily oncomputers.A business manager in a typical managerial situation needs todetermine whether results based on samples can be generalized toa population. Different management situations require differentstatistical techniques to carryout tests regarding the applicability of sample statistics to a population.By the end of this unit you should be able to
Understand the nature of statistical inference
Understand types of statistical inference
Understand the theory behind statistical inference
Apply sampling theory concepts to confidence intervals andestimation
Nature of Statistical Inference
What is Statistical Inference?
We have seen that all managerial/ business situations involvedecision making in situations with incomplete information. Whena particular finding emerges from data analysis the manager askswhether the empirical findings represent the true picture or haveoccurred as a result of sampling accident.Statistical inference is the process where we generalize from sampleresults to a population from which the sample has been drawn.Thus statistical inference is the process where we extend ourknowledge obtained from a random sample which is only a smallpart of the population to the whole population.
Where do we use Statistical Inference?
Let us think of a typical managerial situation: Imagine you are apurchase manager. Your basic problem is to ensure that aconsignment of Aluminum sheets supplied to you by a suppliercorrespond to the required specification of .04 inch thickness.How do you go about ensuring this?One way would be to accept blindly what ever your suppliersclaim. Another option would be to audit each and every item.Clearly this would be both very time consuming and expensiveand would result in an unacceptably low level of productivity.Another option is for the manger to choose a random sample of 100 aluminum and measures them for their thickness. He findsfor example that the sheets in the sample have an average thicknessof .048 inches. On tzhe basis of past experience with the supplierthe manager believes that the sheets come from a population witha standard deviation of .004inches. On the basis of this data hehas to make a decision whether the to accept or reject a consignmentof 10,000 sheets.To solve problems such as this we have to learn how to usecharacteristics of samples to test an assumption about thepopulation from which the sample comes from. This in effect isthe process of statistical inference.The issue facing the manager in the above example is:
Could a sample of 100 aluminum sheets with averagethickness of .048 inches have come from a population with aaverage thickness of .04 inches?
Does the sample estimate of thickness differ from thepopulation estimate due to sampling error or is it becauseour fundamental assumption about the mean thickness of the underlying population is not correct?
Suppose he believes it to be the first case and accepts theconsignment, what is the risk that he runs that theconsignment is flawed and does not conform to qualitystandards of .04 inches?This is just one example of a very typical managerial situationwhere the principles of statistical inference can be put to use tosolving the manager’s dilemma.
Types of Statistical Inference
Broadly statistical inference falls into two major categories:estimation and hypothesis testing. Both are actually two sides othe same coin and can be regarded as representing different aspectsof a technique.Below I briefly explain each of them.
1. Estimation
This is concerned with how we use sample statistics to estimatepopulation parameters. It is not necessary that an estimate bebased on statistical data. All managers make quick estimates basedon incomplete information, gut feel and intuition. Thus an estimateof sales for the next quarter can be based on gut feel or on ananalysis of past sales data for the quarter. Both represent estimates.The difference between a estimate based on intuition and onebased on a random sample is that we can apply the principles of probability allows us to calculate percentage of error variation inan estimate attributable to sampling variation.The sample mean for example can be used as an estimate of thepopulation mean. Similarly the percentage of occurrence in a sampleof an attribute or event can be used to estimate the populationproportion . To explain the concept a little more clearly we canlook at a few examples of estimation:
University departments make estimates of next yearsenrollments on the basis of last years enrollments in thesame courses.
Credit managers make estimates about whether a purchaserwill pay his bills on the basis of past behaviour of customers with similar characteristics or their past repaymentrecord
© Copy Right: Rai University
11.556 101
2. Hypotheses Testing
If we find a difference between two samples, we would like toknow, is this a “real” difference (i.e., is it present in the population)or just a “chance” difference (i.e. it could just be the result of random sampling error).Hypothesis begins with an assumption called a hypothesis thatwe make about a population parameter. We then collect sampledata and calculate sample statistics such as mean, standard deviationto decide how likely it is that our hypothesized populationparameter is correct. Essentially the process involves judgingwhether a difference between a sample and assumed populationvalue is significant or not. The smaller the difference the greaterthe chance that our hypothesized value for the mean is correct.Some examples of real world situations where we might want totest hypotheses:
A random sample of 100 south Indian families finds thatthey consume more of a particular brand of Assam tea perfamily than a random sample of 100 North Indian families.It could be that the observed difference was caused bysampling accident and that there is actually no differencebetween the two populations. However if the results are notcaused by pure sampling fluctuations then we have a case forthe firm to take some further marketing action based onsampling finding.
Colgate Palmolive have decided that a new TV ad campaigncan only be justified if more than 55% of viewers see theads. In this case the company requests a marketing researchcompany to carryout a survey to assess viewership. Theagency comes back with a ad penetration of 50% for arandom sample of 1000. It is now the company’s problemto assess whether the sample viewing proportion isrepresentative of the hypothesized level of viewership thatthe company desires, i.e 55%. Can differences between thetwo proportions be attributed to sampling error or is theads true viewership actually lower.In the next section we shall look at the theory behind statisticalinference. The basis of inference remains the same irrespective of whether the managerial objective is to obtain a point or intervalestimate of a population parameter or to test whether a particularhypotheses is supported by sample data or not.
1.What is statistical inference? What are the different types of inference?2.Why do decision makers measure samples rather than entirepopulation? What are the disadvantages of sampling?
Theory behind Statistical Inference
We now look at the underlying theoretical basis of statisticalinference. The underlying basis of statistical inference is the theoryof sampling distributions.Now we shall briefly review some concepts, which have been dealtwith in more detail in the earlier chapter on sampling.
What is a Sample?
A sample is a representative subset of the underlying population.For each sample that is taken from a population we can calculatevarious sample statistics such as mean and variance. We can takemany such samples from a population and calculate their meanand standard deviations. Given the existence of sampling variationit is likely that there is also going to be some variability in thedifferent estimates of mean and standard deviations. This can beexplained best with the help of an example:Suppose there is a store which sells CDS . We assume it has aregular customer base. A random sample of 100 customers istaken and we find the sample mean age of customers was equalto 42years, with a standard deviation of 5 years. However this isonly one possible sample, which
could have been taken
. A seconddifferent sample may have had a result where mean was equal to45 years and standard deviation of 6 years. To change a sample weneed only change one of the customers. We would expect samplestaken from a population to generate similar if not identical samplemeans. If we take repeated samples such that all possible samplesare taken then we are likely to obtain a sampling distribution of means.
What does this Distribution Look Like?
Logically we can conceive that there is only one sample which willcontain the youngest possible customers and its mean will havethe lowest sample mean. Similarly there will be another couple of samples having the lowest 99 customers. These samples will havemeans, which are slightly higher than the lowest mean. Asomewhat higher number will contain the youngest 98 customersand so on.The majority of the samples will have a cross section of all agegroups and therefore there would be a clustering of sample meansaround what is likely to be the true population mean. Thedistribution of sample means will look like the normaldistribution as shown in the figure1 below.Sampling distribution of sample mean values
Figure 1
This result follows from the Central Limit theorem : if we takerandom samples of size n from a population, the distribution of sample means will approach that of a normal probabilitydistribution. This approximation is closer the larger is n.We do not actually know what form our population distributiontakes: it could be normal or it could be skewed. However it doesn’tmatter, as the sampling distribution will approximate a normaldistribution as long as sufficiently large samples are taken.
Normal Distribution
We now look briefly at some of the key characteristics of thenormal distribution.The normal sampling distribution can be summarized by its twostatistics:
© Copy Right: Rai University
102 11.556
Mean F
standard deviationLogically we can see that the mean of the sampling distributionshould equal the mean of the population. The standard deviationof the sampling distribution is given by /”n, where is thepopulation standard deviation and n is the sample size. Thus thesampling distribution of the mean can be defined in terms of itsmean and standard deviation.However we should be clear We are talking about three differentstatistics:Mean StandardDeviationSample
x s
µ s
Sampling distribution of mean
The three distributions are illustrated below:Sampling distribution of the PopulationThe two distributions are illustrated in figure2 below. As can beseen the sampling distribution of the sample mean is far moreconcentrated than the population distribution. However bothdistributions have the same mean m. 
Figure 2
Application of sampling theory concepts to confidenceintervals.Once we have calculated our sample mean we need toknow where it lies in the sampling distribution of the mean inrelation to the true mean of the sampling distribution or thepopulation mean. It might be higher than the population meanor lower, or it might be identical with the population mean. Whilewe cannot know for certain where the sample mean lies in relationto the population mean we can use probability to assess its likelyposition vis a vis the population mean.From our earlier classes we know that irrespective of the values of and , for a normal probability distribution, the total area underthe normal curve is 1.00. Further specific portions of the normalcurve lie between plus/ minus any given number of standarddeviations from the mean.These results are summarized below:
Approx 68% of all values in a normally distributedpopulation lie within ±1 standard deviation from the mean.Approximately 16% of the area lies on either side of of thepopulation mean lies outside this range. This is illustrated inthe figure 3.
Approx 95.5% of all values in a normally distributedpopulation lie within ± 2 standard deviation from the mean.Approximately 2.25% of area on either side of thepopulation mean lies outside this range. This is illustrated inthe figure4.
Approx 99.7% of all values in a normally distributedpopulation lie within ±3 standard deviation from the mean .This is illustrated in the figure5. Only .15% of the area underthe curve on either side of the mean lies outside this range.
Figure 3,4,5
Standard Normal Distribution
However we rarely need intervals involving only one, two or threestandard deviations. Statistical tables provide areas under thenormal curve that are contained by any number of standarddeviations (plus/ minus ) from the mean. We do this byconstructing the standard normal distribution which isstandardized .Thus all normal distributions with mean andstandard deviation can be transformed into a standard normaldistribution with
=0 and
=1. This transformation is doneusing the z statistic whereZ=
µ/ s
The distribution of the z statistic represents the standard normaldistribution with mean
=0 and standard deviation

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->