Causal studies, General characteristics, The concept of causality...

CHAPTER 6. Causal Research

6.1. General characteristic

As already noted, causal ( causal ) studies are conducted to conclude that there is a causal relationship between events by recording changes occurring in a specially planned marketing experiment. Sometimes this is the only way to establish the existence of the regularities that interest us. We will give three examples showing how diverse the fields of application of causal research are.

Example 6.1

The possibilities of applying the results of causal research in arbitration practice

The well-known corporation "Sportsak" filed a lawsuit by a substantially less well-known corporation, "Kmart"; in connection with the fact that the latter brought to the market bags "Paris Sak". The charge consisted in the fact that this bag looks like one of the bags of the "Sportsak" corporation. To prove this fact, "Sportsman" undertook a marketing study. During the personal street interviews, two groups of women were selected for the experiments. The experiments consisted in demonstrating to them various bags asking them to determine which firm they had released. It turned out that women really consider the manufacturer of light bags of the company "Kmart" company Sportsman & quot ;. Based on these results, the arbitration court demanded from the company Kmart stop the production of bags of this model.

Example 6.2

Opportunities for applying causal research when choosing a promotion policy

Company Eccert wondered if the sales of medicines would increase if they were advertised on the radio in the pharmacy. It was selected 12 pharmacies, which have approximately the same size, geographical location, sales volumes and life span. These pharmacies were randomly divided into two equal-sized groups. All the pharmacies of the first group were equipped with radio broadcasts, and in the pharmacies of the second group, even those radio broadcasts that existed there earlier were removed. Data on sales of drugs in physical and value terms were recorded for three periods: one week prior to the start of the experiment, during the four-week experiment and within a week after the end. It turned out that where radio advertising operates, drug sales are at least twice as high. Having convinced of it, the firm has decided to adjust radio advertising in all of its pharmacies.

Example 6.3

Price policy

The relationship between the size of the discount indicated on coupons was investigated, with the probability of their presentation for the purchase of certain goods: washing powder, toothpaste, etc. It turned out that with the increase in the discount, the likelihood that the goods will be bought by those who previously either did not buy it at all or bought, but rarely. The intensity of the same purchase of goods by those who have previously bought it often does not depend on the size of the discount. Thus, coupons will not help if the reserves for attracting new customers are exhausted.

6.2. The concept of causality in marketing and the conditions of causality

The notion of what is meant by the cause in marketing is somewhat different from how this word is understood in the ordinary sense. Here is a table illustrating this thesis (Table 6.1).

Table 6.1. The meaning implied in the concept of causes in marketing and in the usual sense

Normal Understanding

Marketing Value

The event Y can occur only if X

X is one of several events; in the event of any one of them, an event Y may occur

If an X event occurs, then Y

The occurrence of the X event raises the probability of the occurrence of the event Y

It can be shown that the event Y occurred because there was an event X

We can never strictly prove that the event X was the cause of the occurrence of the event Y.

At best, we can say: experimental data indicate that the absence of a connection between these events is extremely unlikely

Thus, marketing effects usually appear in a probabilistic form.

To conclude that there is a causal link in marketing, you need to verify that the three conditions are met.

1. Is there a joint variation, a joint change in the parameters?

2. Are the events correct in time?

3. Is it possible to influence other factors?

Let's explain the essence of each of these conditions.

Example 6.4

Is there a joint variation, a joint change in the parameters?

Evidence of joint changes can be both a quality and a quantitative character.

An example of qualitative evidence. We can assume that the better the service in the store, the higher the sales in it. This means that in shops with a good service on average, sales should be higher than where the service is bad. If this is not the case, then the hypothesis is not confirmed.

An example of quantitative evidence of joint variation. One study used a random sample of 1000 respondents to study the relationship between education and the purchase of fashion. We divide the respondents into two groups: 500 people with a higher education level and 500 with a relatively low level. The following was found (Table 6.2).

Table 6.2. Research of demand for fashionable clothes, people (%)

Buy fashionable clothes

Respondents with a low level of education

Respondents with a high level of education


322 (64%)

363 (73%)

Not enough

178 (36%)

137 (27%)

Total respondents

500 (100%)

500 (100%)

We see that among respondents with a high level of education, the share of fashion buyers is higher. It seems that education really contributes to the purchase of fashion. But can we conclude that a high level of education leads to an increase in purchases of fashionable clothes? Of course not! One can only say that our observation indicates the existence of a joint variation. As we will show below, this is far from sufficient to conclude that there is a causal relationship.

Example 6.5

Are the events correctly sorted in time?

You need to make sure that the events X and Y occur in the correct sequence, i.e. either first X, then Y, or both. But not at first Y, and then X.

This condition is not always met. For example, a person can make many purchases in a store (X) and have a payment or a discount card of this store (Y). What is the cause, and what is the consequence? After all, it is clear that if a person has a store card, he is drawn to go to this store. And vice versa - if he is comfortable with this store, then it is likely that he will want to receive a card. Therefore, a wide distribution of discount cards does not always and not everywhere leads to an increase in the number of buyers.

Another example. It is noticed that people began to make a decision about buying directly in the store. This is due to the fact that

in stores there was more advertising or, conversely, advertising became more, because the store management noticed changes in the behavior of people?

Consider again the hypothesis that the service in the store leads to increased sales. You have to make sure that you first trained the former or hired new staff and only then sales began to grow. Or, let's say that simultaneously there is a training of the personnel and increase of sales. Both these options do not contradict the hypothesis of the existence of a causal relationship. But it can also be different: the store management, noticing the rise in sales, sent part of the additional revenue to the organization of staff training or hired new well-trained vendors.

Example 6.6

Is the possible impact of other factors excluded?

Note that the related events X and Y can both be the result of the third event. You need to make sure that this is not so. Let's return to the example of analyzing the influence of the level of education on the purchase of fashionable clothes. Here, as we have seen, there is a joint variation with education. And what about the influence of other factors? After all, fashionable clothes are more expensive, and educated people usually earn more! If we break respondents into groups also according to income (Table 6.3), we see that the correlation was false: there is practically no difference in the percentage between the columns of each pair. The illusion was created because of 500 people with a high level of education - 300 relatively wealthy, and out of 500 people with a low level of education - only 200.

Table 6.3. Dependence of the amount of fashionable clothes being purchased on income and education of respondents, people (%)




Respondents' income



Respondents' educational level

Respondents' educational level






171 (57%)

122 (61%)

151 (76%)

241 (80%)

Not enough

129 (43%)

78 (39%)

49 (24%)

59 (20%)

Total respondents

300 (100%)

200 (100%)

200 (100%)

300 (100%)

Now we continue the theme of service and sales. To conclude that a good service increases sales, one must be sure that other factors have not changed simultaneously with the improvement of service: prices, advertising, breadth of supply of goods, quality of goods, competition, etc. There are usually so many factors that, after studying the situation after the fact, one can never say with certainty whether all of them were excluded.

Suppose now that we are still convinced of the existence of a joint variation, the correct sequence of events and the absence of other possible impacts. What else do we need to make a final conclusion about the existence of a causal relationship? In addition to all these evidence, we must have a meaningful internal understanding of why such a link can and should exist. Controlled marketing experiments are conducted to check the three conditions we named and to obtain grounds for internal conviction.

Also We Can Offer!

Other services that we offer

If you don’t see the necessary subject, paper type, or topic in our list of available services and examples, don’t worry! We have a number of other academic disciplines to suit the needs of anyone who visits this website looking for help.

How to ...

We made your life easier with putting together a big number of articles and guidelines on how to plan and write different types of assignments (Essay, Research Paper, Dissertation etc)