Menu ENN Search
Language: English Français

Sample size for random survey in Ethiopia

This question was posted the Assessment forum area and has 1 replies. You can also reply via email – be sure to leave the subject unchanged.

» Post a reply

Toby Gould

RedR

Normal user

16 Jun 2011, 14:11

I have received a question to our Technical Support Service and feel it is better answered by EN-net

Brief project / assignment description : In depth Needs Assessment of out of camp refugees and host population.

We are trying to calculate a good random sample for the assessment, taking into account the following considerations:

- Total population of around 15,000 out of camp refugees
- Total host population as yet unknown
- Refugees are distributed in 8 different locations and several villages within these locations
- We need to survey both the refugee and the host population
- We want to ensure that the sample is representative in terms of location and gender

What should our sample size be and how do we make sure it is stratified?

If you need more information then I can ask the person or put you in direct contact with them

Thank you
Toby Gould

Tamsin Walters

en-net moderator

Forum moderator

22 Jun 2011, 09:42

From Mulugeta WTsadik at UNHCR:

Dear all,
I have got this question in the website,
first of all it is not a real situation in Ethiopia. This scenario is good but I need more clarifications

1. We will identify the areas (8 diff ) what is the distance among the 8 different location, what is the demarcations - river / mountain, or other?.
2. 15,000 refugee are living with the local community? and the are getting the same assistance?

My suggestions as follows: The objective of the survey is both the refugee and the host population.

1st. we know refugee numbers and also needs locating in each refugee HH's in 8 location before sample calculated.
2nd. we need to know a rough estimation ## the local population from the local authorities /health facility /district .., if they do not know the ## of local population, we can count it HH's and also gives a number in each House.
3. Representation is a mater of following the sample procedure during data collection.
4. Other, in where the situation, we used a two stage cluster sampling and also to increasing the number of clusters to address and gives equal chances all locations including to sample unit.
5. If the area do not conduct a survey before, we will used 50% an estimate prevalence (it is understood-not always true depending the population (heterogenous or Homogenous).
Regards, Mulugeta

Back to top

» Post a reply