Aseem Mishra

Yes it needs ethical clearance

Yes the questionnaire is to be verified/certified or validated

The sample calculation is depend on the population

Does starting a community based questionnaire survey requires IRB clearance?

YES

Does the questionnaire is to be verified/certified or validated?

YES, if not previously validated among your study population.

How to calculate the number of participants for the survey?

Check PS Power and Sample Size Software or http://www.raosoft.com/samplesize.html ONLINE

Does starting a community based questionnaire survey requires IRB clearance?

Yes

Does the questionnaire is to be verified/certified or validated?

Yes

How to calculate the number of participants for the survey

You can use the formula z2pq/d2 where p is d prevalence

I agree with the previous responses. Yes, the survey does need to be reviewed and approved by an IRB. To the best of your ability and resoucees, always use survey items that have been validated in previous research. Your sample size depends on your outcomes.

What I would like to add is that it is important to have a clear definition of "community-based." Whether it is truly a communityi-based study or not depends on your sampling frame and methods.

I totally agree with the previous answers. When planning to conduct population survey you need a number of steps:

1. Endorsement of the study itself by some referent body

2. Endorsement and validation of the questionnaire. If you use an already validated questionnaire, then you need only endorsement.

3. Some sort of communication and coordination with key figures in the targeted population to facilitate the survey

4. Definitely you need an estimate of the sample size. For this, you use the formula suggested by Ishag Adam above. It is the simplest formula. From previous studies you need some estimation of P (the proportion of people with a characteristic you study. I do not know what information you gather in your survey but let me assume that one piece of data collected is proportion of smokers among adult males. Lets assume that previous studies put this proportion at .24, The sample size for your study can be calculated as:

Z squared*P(1-P) 1.96 *1.96 *0.24* 0.76

N=---------------------------------= ----------------------------------- = 280 adults at least

e squared 0.05 *0.05

If we assume there is one adult per family, then you need to visit at least 280 families

Note: The sample size will be much bigger of e is smaller, e.g.=0.02 for example N=1750

YES you need ethics clearance.

YES you should use validated scales/questions if at all possible, so you can compare your findings to others in the field

Sample size as per above comments

ADDITIONAL you should pilot test your survey/questions with a selection of people from your target group (e.g. young people, cultural groups, ages) to ensure your language is appropriate and that all questions are understood in the correct context, and the responses you receive actually reflect the questions asked.

You should also seek endorsement from a representative body reflecting your target audience; this may a) help with distribution and b) improve response rates.

I agree with the previous comments, but I suggest the EpiInfo program for sample size estimation (free download), because you have to enter the size of the target population which was not present in the mentioned formula.

Don'r forget to make a pilot study before finalizing the questionnaire.

Definitely you have to test the content validity and reliability of the questionnaire.

Are you doing a community survey? Or, are you collecting a sample from a defined but limited geographic area? The first implies an ecological study where communities are unit of analysis. How do you define this unit of analysis? The sample of this UOA is drawn from a defined population of communities. A community for a case study can be selected from a larger population of communities for its representativess for instance. if communities are defined as US counties a probability sample can be drawn from the total population of US counties. The second approach means that the sample represents a particuliar population characteristic(s). Even if sample is drawn from limited geographic area, the sample is meant to represent a larger population. Omran’s Comments for instance imply that you might be interested in learning if smoking in the defined community sample selected for study has external validity.

Hi, Firstly ethical clearance is a must. Next, you must have conducted your pilot study and based on that you will have better idea of your sample population. TQ

A community-based survey will obtain information but the survey should go through pilot testing for validity, reliability, etc. with IRB clearance. Also it is very important that cultural competency is also assessed so that the survey will yield accurate information.

Very knowledgeable discussion. If you have IRB you have to take clearance. But if there is no IRB from whom we have to take clearance? Validation of questionnaire is very important if you really want to do good work. No of participants can be calculated statistically.

Thanks for sharing a very good question.

Regards

Ethical clearance is a given however, if you are a community based organization or a community researcher that is going to be very difficult to get. If that is the case I suggest you at least try through a local university!.

If you have a sampling frame then you can use epi info to calculate the sample size. If not then you will actually have to create a sampling frame. If the question you are asking is new or has not been asked before then you must test for validity and reliability- sometimes a focus group can help. This is especially important if you need to translate into another language. In that case you have to test the translation and then ask your group to back-translate.

Good Luck

To disagree with the 100 percent agreement that a "survey" requires IRB clearance, it depends upon what you are asking and why. For example, in my field (architecture), a needs assessment survey does not require IRB approval. The way the folks at the UW IRB office explained it to me was: you do not need approval unless you are building theory. If you are just gathering information, no approval is needed.

An IRB review is only necessary to protect the rights and welfare of human research subjects (here is an example IRB Policy from Oregon State U: http://research.oregonstate.edu/irb/frequently-asked-questions/what-institutional-review-board-irb). IRBs must be used when your subjects are identifiable individuals. When subjects are groups or aggregations, such as communities, zip-codes, cities, states, countries, racial or ethnic groups, etc, IRBs are generally not necessary. The exception would be if the size of the aggregations are so small that with a limited amount of data individuals might be identified.

An IRB review is only necessary to protect the rights and welfare of human research subjects. IRBs must be used when your subjects are identifiable individuals. When subjects are groups or aggregations, such as communities, zip-codes, cities, states, countries, racial or ethnic groups, etc, IRBs are generally not necessary. The exception would be if the size of the aggregations are so small that with a limited amount of data individuals might be identified.

It is always best to check with the sponsored research office in terms of IRBs. In many cases the community studies may be exempt but as stated by others they sometimes are needed If the project requires it. We have found that things such as program development and evaluation are often exempt.

we should consider the ethical issues, and clearance.

Ethically approved validated questioner.

Regards

Dr. Kamath Madhusudhana

Dear colleague: all suggestions that you received are correct.

Additionally, I suggest to write all your process. For that objective is useful elaborate a MOP, with your procedures step by step.

Is recommendable use checklists to corroborated if you are ready to start!