Human Research Ethics approval was acquired from the Victoria University Human Research Ethics Committee  (HRETH 028/04) prior to commencement, and all guidelines stipulated within were adhered to. The internet was the chosen medium of the study, as it allows many participants to be reached conveniently and quickly, in comfortable and private settings, helping to reduce social-desirability, and is salient with video gamers as they are shown to be adept internet users .

A website was created introducing and explaining the study, adhering to ethical guidelines as well as guidelines for online research with gamers . These guidelines included the listing of contact details for the researcher, a photo of the researcher, a debriefing page with helpful links, increasing legitimacy through acknowledging the institution overseeing the research, and notification that the results of the research will be made public on the website upon completion. After participants verified their age (18 or over), and agreed to a voluntary consent form, they were taken to the full online survey, which was created using the online survey creation tool, Survey Monkey . To reduce attrition participants were provided with their progress (i.e. how many questions remained) throughout the survey.

To increase the honesty of responses and reduce the response bias of social desirability, participants remained completely anonymous and confidential, with no identifying data collected including email address. Participants were informed of this anonymity on the website and in the voluntary consent form. However, IP addresses were collected to reduce duplicate and false responses. At all possible times, social skills were referred to as ‘communication styles’, to further reduce socially desirable response bias (Riggio & Carney, 2003). The inclusion of a social desirability scale was considered, but as past research has indicated many of the SSI subscales are unaffected by social desirability (Riggio, 1986), as well the anonymous and private nature of internet participation further reducing social desirability, it was deemed unnecessary.

Forums catering to the fullest gamut of video game systems and genres were utilised in order to target the widest range of gamers possible. In the great majority of forums, invitations to participate were posted in the ‘general’ or ‘chat’ section, which were always the most populated, and allowed for a range of gamers to be exposed to the invitation without having to duplicate posts in each different forum sub-section (i.e. Playstation, X-Box, PC, etc.). Any questions posted on forums or emailed to the researchers were answered as fully as possible, in order to develop and maintain a rapport with the community involved in the research.

The data collection system was tested with a pilot study of 30 participants, which confirmed that data collection was successful, and also led to the addition of the verification question ‘work status’. The dataset was then formatted from Microsoft Excel into the Statistical Package for the Social Sciences (SPSS) and analysed. On top of the 621 fully completed participants, a further 250 began the survey but did not complete all questions, and were excluded from the data. Invalid age selections (including blank responses and ages under 18) were also excluded (n = 5), as well as invalid work status selections, including responses of “do not work or study” with another response such as “work full time” (n = 7).

Participants with large unexplained discrepancies between their average daily hours of play, and average weekly hours of play were removed from the data (n = 3). Some participants put in a range of hours in these sections, for example, 15-20 hours of play per week. For these responses a median between the two figures was used. For participants who did not complete the hours of play per week section, but did complete the hours of play per day (with some noting to multiply the previous question by seven), the daily hours of play was multiplied to obtain the weekly figure.

IP addresses were scanned for duplicates to ensure that many surveys did not come from the same IP address in a close proximity of time. Twenty-seven duplicates emerged, the largest of which was five records from one IP address. These duplicates had dispersed start dates and times, they all indicated the same country of residence, many showed different ages and genders, suggesting that the same computer was used by different people legitimately. No duplicates appeared identical in survey responses, therefore no records were removed based on duplicate IP addresses.

In preparation for analysis, reverse-scored items were transformed into new variables using SPSS, and totals calculated for the SSI subscales and other dimensions of the SSI, as well as the RSES and PVP overall scores.