Onderzoeksmethoden 2/het werk/2012-13/Group 2

December holiday gift-receiver associations with students in Nijmegen.

Introduction

Many people celebrate holidays around the end of the year for which they buy presents for other people. We were curious how the relationship with the receiver of these gifts influences different factors like the type of gifts, budgets, etc. In order to do statistical research on this matter there needs to be a questionnaire with which to ask for the various factors that we are interested in. We set out to find out:

Is it possible to create a questionnaire that asks students in Nijmegen what gifts they buy, for whom, during which holiday, around December?

Why 'students in Nijmegen'?

We assume that each category of people who give gifts has its own properties.

For example students and grandparents have different budgets, interests and knowledge. To create a questionnaire that covers every consumer group everywhere is unrealistic because of these different properties. We therefore chose to limit our focus to students.

The reason for focusing on the Netherlands is that we also think that what people give depends on cultural backgrounds. The types of holidays will also differ.

The reason for choosing students in Nijmegen specifically, is partially practical, from an availability standpoint. We have more access to this target audience than to others, which will make our research easier. We cannot however make assumptions based on students in Nijmegen about students, in for example, Amsterdam. Therefore we will only focus on Nijmegen. Other researchers might want to expand on this research and look at more/other groups in Nijmegen, or other places. Comparisons with our results could also be made with students from other countries.

Why are we interested in the different groups of gift receivers?

We also assume that different groups of receivers, receive different gifts. We for example would buy different presents for our grandparents than for our little siblings. We want to research if there is a difference in what our target audience buys for different people in their lives. It might be possible that some people never give gifts to some relationship types.

Our research could show that some groups are very likely to receive gifts of a certain nature, this opens opportunities for more research on these and the reasons why.

Why are we focusing on the period around December holidays?

There are a huge amount of holidays in the world throughout the year. We think that different presents are given for different holidays. So in order to create a questionnaire that is useful will become hard if there is a too big set of possible gifts, dates, etc.

Since we are focussing on students in the Netherlands we have decided to focus on the period around December holidays. We know that in the Netherlands there are two celebrated holidays in December, namely: 'Sinterklaas' and 'Christmass'. Our study might give some interesting insights in the gift buying behaviour of people here depending on the holiday.

This means that the results of our questionnaire might provide opportunities for retailers to gain knowledge about, in this case, students and their gift buying behaviour.

Possible directions

There are a number of directions in which we can go with the eventual questionnaire. We can have a questionnaire that is purely meant to make an inventory of what categories of product are popular for which relationship type. We can however also expand this:

• Where do you buy presents?
  • Do you use online stores?
• When do you buy your presents?
• What kind of budget do you have for presents?
  • Do you save up for buying gifts?
  • How does the recipient influence your budget allowance?
• Why do you buy certain presents?
  • Do you buy what you'd like to receive, or what the receiver would like?

Associations vs actual buying

The title of our research 'December holiday gift-receiver associations with students in Nijmegen.' gives away the fact that we focus on associations that gift-givers have between december holidays and receivers.

This is an important contrast to 'buying behaviour' because this would have implied a study of actual gifts being bought rather than what people might think about buying.

Practical use

As mentioned before, the results of our research could be used for different purposes. The most important would be in advertisement and marketing. Targeted advertisement can be performed during the holiday period.

Christmas gift boxes that employers give to employees are often put together at specialised companies. Our research could provide them with insights on what people like to give to each other, thus what employers might want to give to their employees.

Another example would be financial advice related to budget advice for certain groups, in this case students. Advisors can help students plan for the extra expenses related to holiday gift purchases.

Classifications

The most important product of our research is a classification of the variables that we focussed on. We have made an initial classification for each of them and have then, with the results of our interviews, tweaked this classification to reflect our findings. We have made the distinction between 'simple' and 'complex' paramaters. An example of a simple one is 'gift reciprocation' which has 'yes' and 'no' as possible answers. A complex one is the type of gift, which can have a vast amount of varying possible answers.

Parameters

We will measure the influence of the relationship that the giver has with the receiver on our variables:

• Type of gift
• Quantity: How many gifts does the giver buy for the receiver?
• Joint giving: Is the receiver group of 2 or more people?
  • Who?
• Joint buying: Does the giver buy a gift for the receiver together with someone else?
  • Who?
• Budget
• Idea source: Where does the giver get the idea of what kind of presents to give the receiver?
• Acquisition: Where does the giver buy the gift(s) for the receiver?
• Purchasing time frame: When does the giver buy the gift(s)?
• Reciprocation: Does the receiver also give a present back to the giver?

Method

The end result being a list of possible answers for the questions on a questionnaire, meant that the ideal way for coming up with a classification an interview. An interview has the same style as a questionnaire, namely questions and answers.

Interview

The answers in a questionnaire need to reflect the possible answers people might give for the questions that are asked. We could therefore use questions in our interview that would be similar to those in the questionnaire that we envisioned, with the only difference that the interview is more open, as opposed to a closed questionnaire. In order to limit the answers of the interviewees to the parameters we were interested in however we asked them questions about them directly.

After determining which parameters were interesting to our research we wanted to interview people about each of them in order to get a set of values from which we could distil a classification.

We wanted to find out whether the parameters were influenced by the relationship that the interviewee has with the receiver of the gifts, we therefore needed to ask an interviewee about his gift giving behaviour with more than one receiver. To do this we first needed to find out whom the interviewee gives presents to and then ask questions to determine values for each parameter for each of these receivers. We therefore created a sub-set of questions that we asked for each of the answered receiver-relationship-types.

Pretest

The pretest involved 3 students from the same subject pool as the main study. This pretest allowed us to test whether the responses matched our expectations and see whether they would be useful to us in the classification process. After tweaking the interview questions we came up with a final version.

The final interview we created (also see Appendix): The Interview

Selection and size of test group

We have chosen to work with a test group size of 8. The first 3 interviews were conducted in the pretest to fine tune our interview and our recording process, the remaining five were used in the classifying of our parameters.

Due to limitations in time and resources we have made our, hand picked, selection of interviewees from our own social environment. The result of this has been that we have a very international group of, non-random, interviewees.

Recording

All the interviews have been recorded in order to retain all information that we received from the interviewees and as proof for our results.

The recording was performed using laptops in mostly quiet environments. The location was often determined by the availability of our interviewees.

Transcription

Transcription has been performed twice for each interview in order to guarantee the correctness of the transcriptions.

No special software was used to perform the transcriptions. And typically took between 15-20 minutes.

Tagging

Once the interview recordings were transcribed we tagged each of them for the parameters that we were interested in using Microsoft Word's 'comment' feature:

Classifying

We used the tagged files to update our initial classification by looking at every parameter and taking inventory of the answers that were given to the related questions.

We were able to come up with classification based on the answers and then verified them by going through each interview and checking whether the answers fit our classifications. By performing multiple iterations were able to tweak final versions for each parameter.

Example

An example of this fine-tuning occurred with the gift-type parameter.

After the first pass of our tagged files we classified this as follows:

• Emotional Gift (Photos, Written poem, hand-made things…)
• Experience gift (Concerts, Bungee jumping, Travel, etc…)
• Aesthetic (Decoration, etc)
• Animal
• Food and drinks
• Sports
• Toys
• Clothes and accessories
• Electronical Media & Video games
• Gift card
• Beauty and Care
• Practical

This yielded too many answers however to be practical in a questionnaire. We therefore analysed the possible meanings that people could attribute to each category. For example 'toys' and 'sports'. Under sports you could count tickets for a football game, this however could also fall under 'experience'. So 'sports' could become 'sport equipment'. Then again though for an adult a football is 'sport equipment' for a child it is a 'toy'. We therefore grouped 'sport and toys'.

The final classification for gift type became:

• Emotional Gift (Photos, Written poem, hand-made things, animals…)
• Experience gift (Concerts, Bungee jumping, Travel, etc…) ‘going places’
• Aesthetic (Decoration, etc)
• Beauty and Care (clothes, jewelry, perfume, etc…)
• Electronic Media & Video games
• Food and drinks
• Books
• Sports & toys & gadgets
• Gift cards & Subscriptions (Gift card, newspaper subscription, spotify, etc…)
• Practical gifts (Tools, Cooking utensils, ‘things for the house’, etc…)

Processing

We created a spreadsheet in which we listed all the parameters and then went through the interviews. Using the created classification and tags we could quickly associate the relevant parameters and their values. During this process we also made changes to our classification.

For example; we added 'books' as a gift-type. A more drastic change occurred in our classification for 'purchasing timeframe'. The initial classification for this was:

• 0-2 days
• 1-2 weeks
• 3-4 weeks
• > 1 Month

Obviously there is a gap here between 2 days and 1 week. Filling it up by simple adding '2 days - 1 week' would not however solve the issue. The problem is that 0-2 days represents 'last minute' '2 days - 1 week' feels very clumsy. In order to fill the gap and retain the meaning of a very small amount of days we adjusted the classification to:

• after
• 0-3 days
• 4-10 days
• 10-20 days
• 20-30 days
• >= 30 days

Results

Classification

Our research has proven very effective in creating a, what we believe to be, useful classification of our parameters.

The chosen parameters were:

• Relationship type - Giver-Receiver relationship type.
• Type of gift
• Quantity - How many gifts
• Reciprocation - Receiver gives the giver a gift as well
• Budget
• Purchase time fame - When does the giver purchase his gifts?
• Idea source - How does the giver come up with an idea for a present?
• Acquisition location - Where does the giver get the gift from?
• Joint buying - Does the giver purchase the gift together with someone else?
  • With who?
• Joint giving - Does the giver purchase one gift for more than 1 person?
  • For whom?

Simple

A few of these parameters are simple and have a very simple classification.

• Reciprocation: yes/no
• Quantity: [0, 1, >= 2]
• Joint buying: yes/no
  • With whom? This will follow the classification of relationship type.
• Joint giving: yes/no
  • Who? This will follow the classification of relationship type.

Complex

The following parameters were of a more complex nature.

Budget

• >200
• <= 200
• <=150
• <=100
• <=50
• <=20
• No specific budget

Purchasing time frame

• after
• 0-3 days
• 4-10 days
• 10-20 days
• 20-30 days
• >= 30 days

Relationship type

• Father
• Mother
• Parents
• Brother
• Sister
• Partner
• Friend
• Nephew
• Grandfather
• Grandmother
• Colleagues

Type of gift

• Emotional Gift (Photos, Written poem, hand-made things, animals…)
• Experience gift (Concerts, Bungee jumping, Travel, etc…) ‘going places’
• Aesthetic (Decoration, etc)
• Beauty and Care (clothes, jewelry, perfume, etc…)
• Electronic Media & Video games
• Food and drinks
• Books
• Sports & toys & gadgets
• Gift cards & Subscriptions (Gift card, newspaper subscription, spotify, etc…)
• Practical gifts (Tools, Cooking utensils, ‘things for the house’, etc…)

Acquisition location

• Self made
• Online Specific
• Online Search
• Physical Specific
• Physical Browse

This is an interesting classification as it shows how people discover presents as well. 'Physical Browse' and 'Online Search' associations tell us that people are less sure about what to get for a certain relation. The distinction between online and physical shows us differences in types of gifts too however. Matching type of gift to acquisition location might be an interesting statistic in its own right.

Adding gender of the respondents here might also prove interesting for marketing purposes.

Idea source

• Physical Browse
• Internet Search
• Suggestion from others (friends, family, etc)
• Asking receiver
• Customary
• Magazines, Newspapers, etc
• Interests and hobbies of receiver

Transcriptions

The transcriptions

Statistics

Though the data we gathered will not have a very solid statistical meaning, we still found some of them interesting to display as a demonstration of the classification.

Type of gift

When put into a graph the type of gift in relation to the relationship type between the giver and the receiver looks like so:

We can see that Brothers are associated with the most diverse types of gifts. We can also see that gifts that are specifically associated with one parent are different from the types of gifts that are associated with two parents together. Stereotypically father are associated with practical gifts.

Budget

Statistics on the budget are less useful as there is a clear need for a more statistically elegant test group size. In general it can be seen that Fathers are related with bigger budgets.

Purchase location

Interestingly offline locations are still a popular source for purchasing gifts, even when people know exactly what to get.

Improvements for future interviews

Recording

Though the quality of all the recordings was OK, generally one might want to have more professional recording equipment and quieter environments to perform interview with and in. We discovered tiny mistakes in the transcription process that would have been prevented by working in a better environment.

Test group composition

Due to limitations in time and resources we have made our selection of interviewees from our own social environment. The result of this has been that we have a very international group of interviewees. This is not a problem when you look at the initial problem statement 'students in Nijmegen', but the results could be different when the research is performed again with exclusively Dutch students, as the problem statement suggests.

December holidays do not apply everywhere

Something that we really did not expect, despite the composition of our research team, was that Chinese students don't give presents during December or anywhere close at all. One of our respondents had trouble understanding what we meant and turned out to be fabricating answers at some point. We were forced to dismiss the results from this interview. Having a more origin-consistent subject pool in revisiting this subject will prove to provide better results. In hindsight it might have been good to not define the holiday period as 'December', but rather ask about the subjects personal gift-giving celebration habits. The results for presents given at easter would be equally interesting if we were able to show that this holiday is of equal value to the subject as Christmas is to others.

Rather than holidays one might also look at birthdays. Something that does come into play here is that some cultures don't celebrate birthdays in the same way either. Future research might be needed to see whether the level of importance of gift-giving related events changes the findings. Luckily, our classification can also be used for this, be it with slight modifications.

Conclusion

We are very happy with our final classification of the parameters and the way in which we got them. There are a few thoughts that we'd like to expand.

Getting some the classifications were harder than others. 'Idea source' is a parameter that is not very obvious from the beginning like some of the other parameters. It is fairly easy to predict budgets, even without doing interviews, but where people get their ideas is a lot less straightforward. We also believe that our subject pool has influenced the classification. We think that when creating the eventual questionnaire further research should be done concerning the classification of this.

Doing pretests allowed us to fine tune our results and show us where interesting data was to be found. We came up with the 'quantity' parameter and when to ask about it during the pretest.

An element of 'gift type' that we are not 100% sure about is the 'Electronic Media & Video games' category. It might be worthwhile to split this up in 'Music & Film' and 'Video Games'. Pretests when launching the questionnaire, might be able to bring up a final decision here, this will also be influenced by the final purpose of the questionnaire.

The general approach has been good though, we liked the ease at which we were able to create classifications for previously unknown parameters. Something that has skewed our results has been the composition of our subject pool. To get better, more generally applicable results, one would have to interview a more diverse set of subjects than we were able to do.

The classifications as they are though are applicable in a 'production' ready questionnaire however. There are no parameters for which we believe it would not be valuable to have them answered in the presented classification.

Appendix

• The Interview
• The transcriptions

Notes

And earlier version of this document has been archived: Archive