Menu ENN Search
Language: English Français

Sample size for IYCF assessment based on Step-by-step guide

This question was posted the Assessment forum area and has 8 replies. You can also reply via email – be sure to leave the subject unchanged.

» Post a reply

Anonymous 81

Public Health Nutritionist

Normal user

26 Jun 2010, 15:49

My question is with regard to determining the sample size for conducting one-time an independent IYCF assessment for children 0-23 months. According to the step-by-step guide, the total sample size is determined first by calculating the sample size of each IYCF indicators (10) and then the largest one (from indicators with wide age ranges) will be considered. Then this number will be multiply by four (as there are four 6-month age intervals between 0-23 months). Let's say if the largest number is 330, the sample size will be 1320 children 0-23 months. If the assessment is cluster sampling which is the commonest, it will be 2640 children 0-23 months (assume DE is 2). It seems to me that this number (2640) is too much in terms of time and resource especially if the survey area is area is sparsely populated, mountainous and harsh weather condition. Of course, i can understand the technical justification behind but it is not easy to operationalize . We are planning to conduct IYCF assessment in one district but according to the guide we need to visit more than 2500 infant and young child just in one district . So, is there a solution to compromise/adjust the number?

Title of the guide is:
Infant and Young Child Feeding Practices: Collecting and Using Data: A Step-by-Step Guide. Cooperative for Assistance and Relief Everywhere, Inc, CARE 2010).

Marie McGrath

ENN

Forum moderator

30 Jun 2010, 09:18

Dear Kiross,
See below for a response from Kirk Dearden, author on the Step-by-Step Guide:

A sample of 2640 is almost certainly far too large and not feasible given NGOs' limited resources. An initial calculation of 330 per 6-month age group is probably too large and the design effect may be too big as well. 1.5 might work for DE. I would suggest re-running the calculations with these issues in mind. If the resulting sample size is still quite large, our colleagues may simply want to do a 30 x 30 cluster survey of children < 2 years of age--keeping in mind all of the limitations of this design which are mentioned in the Guide.

Kirk is willing to engage in conversation on this issue, should a reader(s) have additional questions.

Best regards, Marie

Anonymous 425

Normal user

6 Jul 2010, 17:35

Dear Marie,

I thank you for your reply. For furhter clarification, i will contact directly with kirk Dearden

Ranjith

Normal user

24 Sep 2010, 16:30

Where can I get a hard copy of the "Infant and Young Child Feeding Practices: Collecting and Using Data: A Step-by-Step Guide. Cooperative for Assistance and Relief Everywhere, Inc, CARE 2010)"?

Marie McGrath

ENN

Forum moderator

24 Sep 2010, 16:54

Dear Anon, If you send your postal details to the ENN, I can send you a hard copy as I have a few at the office. Email your postal details to: marie@ennonline.net

Ranjith

Normal user

3 Feb 2011, 16:53

I'd like to determine the sample size for a KAP baseline survey. One of the indicators in the survey is assessing the care seeking behaviour when a child is sick. There will be a KAP endline survey after 1 year.

Can I use the excell sheet that comes with this guide to calculate the sample size for this KAP baseline (since the fomula is the same?)?

Mark Myatt

Consultant Epidemiologist, Brixton Health

Frequent user

4 Feb 2011, 10:17

See: http://www.en-net.org/question/338.aspx This addresses sample sizes and the weaknesses of the whole KAP survey idea. Survey scientists can be a childish bunch ... they often call KAP surveys "KRAP surveys" because the output is usually crap (defined as "the husk of grain; chaff; something of poor quality; something that is rubbish; nonsense; faeces; a euphemism for shit; useless object(s)" - wiktionary).

Everlyn Matiri

Normal user

5 Dec 2017, 11:48

Hi Matt,

I just wanted to understand your response regarding KAP surveys. Does it mean doing a KAP survey is KRAP (crap) as stated? If this is the case, why? and what options are available?

Mark Myatt

Consultant Epidemiologist, Brixton Health

Frequent user

5 Dec 2017, 12:48


See this thread on EN-NET

http://www.en-net.org/question/338.aspx

The link above does not seem to work?

Back to top

» Post a reply