Question

Topic: Student Questions

A Survey Question Without A Meaning ?

Posted by Anonymous on 6/4/2005 at 7:37 AM ET 50 Points
In a survey i am analysing as an assignment in school, we have this theatre.
a question in its survey states:
" The prices in this theatre are as with competing theatres" ..
Now... i feel that this is aompletely useless question as it really doesent measure anything... am i correct?

the background for this survey is that the theatre is worried it might not create enough revenues the coming year. They have therefore undertaken a survey, sendt out to 200 client, whereof 70 was returned...

One of the questions asked in that survey was the one above...

please post your comments....
To continue reading this question and the solution, sign up ... it's free!
SUBSCRIBE
Or, SIGN IN

RESPONSES

  • Posted by wnelson on 6/4/2005 at 8:44 AM Accepted
    I'm assuming that the survey is written on a multipoint scale from strongly agree to strongly disagree. The survey question is obviously trying to test a hypothesis that the theatre revenues might decline because the price is higher than the competitive venues. But, it is written poorly such that only if the answer is "strong agreement" will you get good information. Strong agreement would mean that prices are in line with other theatres and so if revenues decline, it is most likely due to another cause. If, however, the results of this question are "disagree" - you won't know what this means - that the prices are too high or the prices are too low. Therefore, the question, as written, may yield a vague answer, and could be useless.

    In a case where sample size is small and tallying the results does not have to be statistically pure, deviating from the multipoint system would be reasonable and this question could have been worded:

    How do the prices at this theatre compare with prices of other theatres: A) prices are higher than other theatres; B) Prices are the same as other theatres; C) Prices are lower than other theatres.

    Another question that might be necessary to understand if effect of price on revenue for this survey is:

    When making a decision between theatres, A) I choose the lowest price; B) price doesn't matter - I select the theatre based on other factors; C) I choose the highest price.

    I don't know about you, but I choose a theatre based on the what is showing there, first. And I choose tickets based on the location, independent of price, where thre are multiple prices based on seating (like for a concert). For a movie theatre, where I may choose from various places with the same show, I choose on location, cleanliness, and price with all other factors the same. Also, for some luxury goods, the price-demand curve is reversed. Customers look for the highest price versus the lowest price. Theatres might fall into that category.

    Of course, an even better approach would be, in conjunction with the survey, for someone in the theater to drive around the city to other theaters and look at the box offices to SEE how the prices compare. You might think that if you did this, you wouldn't need the question. But, remember that consumer choices are based on perceptions. If the customers perceive the prices to be higher - even if in reality they aren't - then they will behave like the prices are higher. Perception is reality. So understanding that the prices are perceives as higher or lower than competition when in reality they aren't is important.

    I hope this helps answer your question.

    Wayde
  • Posted by wnelson on 6/4/2005 at 2:33 PM Member
    Rubukas,
    OK, I understand this a little better now. I believe the analysis you are doing - crosstabs, regression, correlation - are adequate for this problem. Using Visit as a "dependent variable," then if the analsis shows that there is a positive slope parameter and significant significance for the regression, then you can conclude that people visit the theatre more if the price is closer to the competitive price. If this is that case, the action is easy. Make price the same as competition. If the slope parameter is negative - meaning that more different the customer believes the price is, the more they visit, then because of the way the question is presented, you can't decide on an action - make prices higher or lower than competition. If there is no significance, it means price doesn't matter.

    I believe this is probably the intent of the problem - to show that you should make price the same and people will visit more frequently. Different, they won't. The common thinking is that higher price will drive less revenue and if you make this assumption, the question is probably adequate.

    Wayde
  • Posted by wnelson on 6/4/2005 at 6:45 PM Member
    Rubukas,
    I am familiar with SPSS, though I haven't used it and I don't have the s/w. I've gone through a tutorial and the program and output is pretty intuitive. If your data file is in the form of an excel spreadsheet, then I can look at it for you. Please feel free to send it. See my eMail address in my profile.

    But, as you are doing that, please tell me...is the correlation between VISIT and PRICE significant? If this is the case, and if the slope parameter is POSITIVE, this means that people visit more frequently when the price is close to competition. Your recommendation, therefore, is "make PRICE close to competition." Stated this way, it's irrelavant if the price is high or low. Just make it the same. Right?

    If on the other hand, the correlation is not significant, then VISIT is not tied to PRICE. Thus, you don't have to do anything with price.

    If you have correlation and the slope iparamerter is NEGATIVE, then this says that you need to make price very different from competitors to improve VISITS. For this case, you don't know if your action should be "raise price" or "lower price." The only recommendation here is to survey again with a question to decide if a higher price or a lower price is better. I am betting your profs didn't give you this case and that PRICE is significantly correlated to VISIT and the slope parameter in the regression is positive.

    Bottom line: It's a bad survey question and I believe that staring at the data and analysis isn't going to provide any other meaning. We're only going to have two options depending on the data:

    1) Correlation is insignificant. ACTION: Ignore price - it doesn't matter
    2) Correlation is significant, slope parameter in the reguession equation is POSITIVE. ACTION: Keep price the same as competition.
    3) Correlation is significant and the slope parameter in the regression equation is NEGATIVE. ACTION: New survey needed to know if the right action is to raise or lower price.

    As I said above, if you have the SPSS output in Excel form, I can look at it and see if I can see anything more. Send away!

    Wayde
  • Posted by wnelson on 6/4/2005 at 7:08 PM Member
    Rubukas,

    I want to be more specific about this issue with PRICE. You and I are thinking about testing our PRICE. The problem is set up to test the CLOSENESS of our price to the competitive price- which is the difference of our theatre's price and the competitors' price. So the action we can take as a result of the data is to make our price CLOSER, or make them NOT CLOSER. If it's make it CLOSER, then it doesn't matter if our price is too high or too low...the action is to adjust price to be closer - or the same is the best. For this case, the analysis would show correlation is significant and the regression equation slope parameter would be positive. It should be adequate for your action to make the price the same as the competitors' price. Given this question style, that's all you can conclude.
  • Posted by Chris Blackman on 6/5/2005 at 5:17 AM Member
    Actually the question (right or wrong) is not just about the closeness of the price to a competing theatre's price.

    It's really about the respondent's PERCEPTION of whether the price is close or not.

    So if you are saying the analysis indicates people don't change their visitation habits with regard to their perception of price-competitiveness, then what you are diagnosing is that price is not a critical factor.

    Do the responses indicate there is much perception by respondents of better/worse price competitiveness in other theatres?

    So a better question for the survey might have been:

    What factors drove your choice of this theatre for your visit?
    - Location
    - Ambience/Decor
    - Car Parking
    - Proximity to Public Transport Routes
    - Proximity to other entertainment destinations, e.g bars, clubs or restaurants
    - The movie program showing at the time

    Hope this helps
  • Posted by wnelson on 6/5/2005 at 2:51 PM Member
    Rubukas,
    Let me give you a description of the process to explain the answer to your question concerning the comment on "too few observations." The methodology - crosstab (short for crosstabulaton) is a methodology of organizing the data from two columns of data and tabulating the frequencies for each cell-pair. In this problem, one column is VISIT and another is PRICE. The analysis of the data to determine if there is a significant relationship between the two factors is the Chi-square test. The test compares the actual observations to the expected observations for each cell. An issue with the Chi-sqaured test in this application is that it doesn't work well for small expected values. When the expected value for a particular cell is less than, say, 5, then the significance of the results is not sufficient. This is the point of your previous professor's note. Your total sample size is more than adequate. It's the expected values that need to be reviewed.

    In your data, for many of the crosstab cells, the expected value is much less than 5. In a case like this, traditional multiple regression is a viable alternative. Try doing a multiple regression on this data.
  • Posted by Chris Blackman on 6/6/2005 at 8:42 AM Member
    Rubukas

    I student problems I always encourage viable assumptions as long as they are declared.

    Professors aren't looking for the one right answer. They are looking to see how you thought aboiut the answer. Making and declaring assumptions in the face of a shortage of information involves thinking the problem through.

    I remember an engineering question once: "Calculate the efficiency of a windmill". That was it. A whole raft of assumptions and declarations had to be made before the calculation could be shown.

    Hope this helps.
BACK TO TOP

Post a Comment

Related Discussions

Community Info

Top 25 Experts
(Student Questions)