It should also be mentioned that representative examinations are only valid if the user (the subject) regularly uses at least one Internet service.
In classical representative surveys, the participants in the examination are addressed directly in the respective institute (= external selection) and in the case of www surveys the participant decides himself (= self-selection) whether he wants to participate in the survey. Distortions in the sample composition are therefore unavoidable.
An influencing also takes place through the selection of the pages on which the reference to the surveys “lights up”. Again, distortions of the sample can take place.
Socio-demographic composition of Internet users
Unfortunately, research on this is rarely done, the reason being that there is a low prevalence in the individual countries. It would need in Germany e.g. 6000 screening interviews to survey 1000 representative selected users, which is associated with high costs! Most of the studies that have been carried out are projects whose results are not published. The exception is “Online Monitor” and the Essner research institute “Academic Data”. Meanwhile, more will be determined! At the time, “Online Monitor” found that 69% of men and 31% of women are represented on the Internet. Furthermore, large differences in the professional status of Internet users have been noted. It should also be noted that this occupational position and also the socio-demographic composition differ greatly from the total population: 11% of the users were between 50 and 59 years old in 1999, with the proportion in the total population being 23%. At the time, 31% of users were executives, but the proportion of those in the total population was only 24%. At the time of the study, these were “over-represented” and officials were under-represented, with 7% of the public using the Internet and 9% of the total population.
Net income also shows a similar picture, but at the time it makes sense to look at the former costs of the internet!
Finally, it can be said that high earners and those with higher positions were more represented on the Internet. However, this has changed drastically in the last 10 years! You do not need much money to get a computer with an internet connection. Meanwhile, the computer has almost a status like the television.
Questionnaire studies
In the following, we will now describe and explain the various possibilities of questionnaire surveys via the Internet in more detail. What advantages and disadvantages there are in the various possibilities of the surveys and we should approach the investigation and also, which rules are to be observed.
Service-specific examination procedures
First of all, it should be clarified which elements should be included in a questionnaire.
Namely:
the name of the investigator a clear and meaningful examination title statements on the meaning and purpose of the investigation the probable processing time
a note if the participants will receive a result report.
There are three ways to distribute a questionnaire over the Internet
1. Surveys by e-mail
2. Surveys in newsgroups
3. Surveys in the WWW / online questionnaire
Surveys made by e-mail or in newsgroups are predominantly text-based and there is asynchronous communication between the investigator and the subjects. However, online surveys can create a certain degree of human-machine interaction through the large multimedia possibilities and the use of other application programs.
Which method of questionnaire distribution is used depends entirely on the investigator and his computer skills.
During our research on the Internet, we noticed that the trend is more and more to questionnaires in the WWW so online questionnaires goes. Therefore, we will go into all the possibilities with the advantages and disadvantages, but especially with the surveys in the WWW with an example to show why this possibility is certainly currently the most widely used.
Surveys by e-mail
In this method, the questionnaire is sent directly to the mail, usually as an attachment to the participants.
On the one hand, the advantages of this type are the high response rates, on the other hand, modern e-mail programs offer the possibility of handling very large address directories. It is important that: the mail merge function of the e-mail program is activated, as otherwise all addressed addresses are visible to the participants. First, this leads to the problem that the mail is illegible and secondly, this is also questionable under data protection law.
Another advantage of the surveys of e-mail is that it is possible for remailers to allow anonymous participation in the investigation.
In a nutshell, this remail works so that the user sends their email to a remailer. This removes all data from the e-mail, which could allow conclusions about the sender, and forwards them to another remailer. Only after the email has passed through a predefined number of remailers (it should be at least 2 remailers), it finally ends up in the inbox of the end recipient.
By programs such as Mixmaster or Quicksilver allows people to send anonymous emails or post anonymously to newsgroups.
Unfortunately, these remailers are often used for the wrong purposes and are often criticized. Especially through mass emails, anonymous propaganda campaigns or unwanted commercial advertising messages, the so-called SPAMS, the continued existence of such remailers is endangered.
However, it is very useful for sending questionnaires, as some subjects, especially on tricky subjects, may be more likely to respond if they know it is anonymous.
However, sending the questionnaires via e-mail also has some disadvantages. Namely it is not possible to attach complete questionnaire programs or test programs to the e-mail as this can lead to very long loading times among the participants. This can lead to a systematic exclusion of participants who do not have the appropriate software.
In addition, as mentioned above, the mail merge feature should not be turned off due to privacy issues and the readability of the e-mail.
To ensure that the questionnaire is presented equally well to all subjects, compliance with the following technical standards has proven itself:
Questionnaires should be created in ASCII character set. In doing so, umlauts are broken down, ie ä in ae, for example.
A line should not contain more than 65 characters, otherwise an unwanted line break may occur at the receiver.
The e-mail should not contain any formatting instructions, such as nothing printed in bold or no tabs.
It should be made clear to the person writing off, where you got your e-mail address from.