Another problem you have to consider is about respondent’s willingness to participate. You have to have a reliable database of lead contacts and be aware that response rate is very low, commonly around 5% to 10% so if you need a sample of 400 subjects, for example, you have to contact at least 8000 people.

Of course, never forget the main characteristics of your sample otherwise your results will be biased.

Hi. You should find this article useful for your justification. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3884902/

Also, this previous discussion on similar issue: https://www.researchgate.net/post/Is_the_google_forms_is_the_good_for_survey_method_to_collect_primary_data_on_E-commerce

Hi,

There are a lot of pros and some obvious cons of web surveys:

Pros

1) cheaper and easy to develop. Advice though is try avoid the free platforms tempting as they may be. They can often get filtered into spam

2) easy to administer

3) data often automatically entered into database or easy to upload. This reduces human error

4) quick analysis time resulting in quicker reporting time too

Cons

1) Need to be more aware of data security and privacy

2) Risk of missing certain sectors in the community esp where internet access is low (rural, low income), elderly

This is a good resource:

Internet, phone, mail, and mixed-mode surveys : the tailored design method

Dillman, Don A. ; Smyth, Jolene D. ; Christian, Leah Melani

ISBN: 9781118456149

4th edition.

Wiley, 2014

Dear Milad Mehdizadeh,

In my opinion Internet (Web)-based survey has many advantages.

The advantages for instance, you will get a wide-range of research participants, easier to administer the survey and organize the data, and you will save a lot of cost and time, etc.

nonetheless there are some disadvantages we need to anticipate.

such as, sample that comes from outside the targeted populations, research participants that not intended to take the survey, extreme data from "unwilling-to-participate-people" that tend to answer randomly or extreme, missing data, etc.

several things you need to prepare, such as:

1. A clear instructions about the survey

2. Make sure you take samples from the targeted populations

3. Some foreword or descriptions about the survey, and a statement from researcher that expect the research participants take the survey seriously.

4. etc.

Overall, Internet (Web)-based survey are reliable, but you need to prepare several things to make sure that the survey is working well and prevent all factors that can affect your data.

Best Regards,

Riangga Novrianto

I agree with all the prior comments and I am a big fan of web-based surveys. However, with some clients where trust in leadership is low, web-based surveys are viewed with suspicion resulting in possibly false answers to conceal their identity. With web surveys, it's important to make clear how you will protect confidentiality of responses.

Another problem you have to consider is about respondent’s willingness to participate. You have to have a reliable database of lead contacts and be aware that response rate is very low, commonly around 5% to 10% so if you need a sample of 400 subjects, for example, you have to contact at least 8000 people.

Of course, never forget the main characteristics of your sample otherwise your results will be biased.

I think that depends on the kind of research you are conducting. But one of the main advantages is the amount of people you could reach with less effort. In order to get more realiability on the answers, you can create questions that may help you with this control. Another possibility is to randomize the questions, to reduce the bias on the answers.

Some good answers.

One of the themes that keeps coming up is the reliability of the data. There are a couple of ways to address this.

1) screening question and logic. A lot of online software allows you to create a journey through the survey based on the answer to a question so only relevant questions to the individual are ask. If they don't pass the screening question it can take them straight to the thank you page.

2) Like other methods data cleaning is important. There are several tools one can use which clean out "dirty" or "rogue" data i.e. straight liners and patterns within multi choice questions, more advance ones can also identify inconsistent responses.

3) Some people like to stack a survey by doing it multiple times. This can also be cleaned up by captureing meta data IPS address etc. This takes a bit of human input as soemtimes you may get 3 or 4 results from the same address but when you check the demographic data this is different so the assumption is the survey has been done by different people in the same household