Data on restaurant tipping from a Norwegian survey experiment

The data are related to the research article “Peer effects on restaurant tipping in Norway: An experimental approach [1]”. The data stem from a survey experiment performed on students attending the Inland Norway University of Applied Sciences in December 2019 and pertain to 709 observations. The key variables are (1) tip amount, (2) peer tipping, (3) bill size, and (4) service quality. The data also contain information on background variables such as previous experience from the service industry, financial situation of respondents, gender, and age. We display the raw data organized in such way that they can be easily downloaded and used directly to either (1) replicate the analyses performed in the related research article, or (2) to run one's own analyses on the topic of interest. The data may also be useful to lecturers teaching students about the key concepts of survey experiments and causal modelling.


Specifications
Social Sciences (General) Specific subject area Business Economics with a special focus on the restaurant industry and tipping practices. Type of data

Value of the Data
• Data on tipping behaviour traditionally stem from regular surveys, exit-surveys, or transaction data from the restaurant. The data offered here use an experimental design and contain information on tip amount for various levels of service quality, bill size, and peer tip amount. • Academic researchers and analysts working in the restaurant/service industry can benefit from these data to get new insight on what causes variation in tipping behaviour and particularly to examine the effects from peers tipping behaviour. Lecturers may also benefit from using these data when teaching students the key concepts of survey experiments and causal modelling in general. • The data may be used to gain further insight by replicating and extending the analyses performed and reported in the related research article. For example, it is possible to include more explanatory variables, transform some of those already used into new variables, create interaction-variables, and more. The data and survey experiment in general may also be used as a starting point when other scholars want to create their own experiments. • The data are collected in Norway where diners are recognized by a high stiffing probability.
This makes the data unique in a tipping context and adds additional value.

Data Description
The data stem from a survey experiment carried out among students of a large Norwegian university college. The questionnaire is provided as a supplementary file. In Table 1 we include information about the various items included in the data table accompanying this article. The table includes information about the values the various variables can assume and the mean/standard deviation. Please note that we use a "-" to indicate that the mean and standard error do not make sense.

Experimental Design, Materials and Methods
The data stem from a survey experiment (see [2] and [3] ) carried out on students at Inland Norway University of Applied Sciences and the questionnaire is provided as a supplementary  [4] The item Questionnaire from the data description refers to the questionnaire version the respondent was given, in which the items service (ok/very good), bill size (310/510), and Peer Tips (0/30/60) were manipulated in the survey instrument. A random number generator assigned each respondent with one out of the 12 different versions of the questionnaire where the variables service (ok/very good), bill size (310/510), and Peer Tips (0/30/60) assumed 12 unique combinations for each of the 12 questionnaire types (2 × 2 × 3 = 12). The values the three experimental variables took on for each questionnaire type may be read directly from the data file and are summarized in Table 2 . The table shows that each level of the various items occurs in a certain fraction of all the questionnaires. For the items service and bill size the two levels (ok/very good and 310/510) occur in 6 out of 12 questionnaires each. The levels of the item Peer Tip (0/30/60) occur in 4 out of 12 questionnaires. The way the manipulation is set up ensures that all combinations of the three key items ( service, bill size, peer tips ) are covered.  The data file accompanying this article consists of the raw coded data points as they were provided by the respondents in the survey experiment. For the benefit of those wanting to replicate the analyses presented in the related research article we have calculated the variables needed to do so from the raw data. These are: (1) Tip in % of Bill Size ( T ip Amount/Bill size ) ,

Ethics Statement
All respondents participating in the survey experiment provided their informed consent before participating.

Declaration of Competing Interest
None.

Supplementary Materials
Supplementary material associated with this article can be found in the online version at doi: 10.1016/j.dib.2020.106441 .