The best way is randomization. For example every third person in an object for 2 hours or something like that

One easier way would be the Slovin formula

N/(N*(d)^2+1),

with N=Number of Population

d=1-degree of confidence

For example, we have a population of 29,000 and we desire degree of confidence of 95%, then 29000/((29000*(0.05)^2)+1), would be the minimum sample of 395 respondent.

It's not the best way of course, but should be sufficient for your purpose.

Hi Imam Salehudin, could you send me journal articles that support its use? Thank you.

I can deduce that there is not a random frame. Either a way to apply a sistematic routine to select an element. It is better to redefine the population.

The formula you need to use is :

n = p (1 - p) / e ^ 2

where n is the size of your sample

p = proportion you study

1 - p = opposite

e : precision needed 1% , 0.1% .... and so on

hope that it will help !

Patrick

Dear Hiram,

In the formula of Patrick you need to include your level of confidence.

I.e. for a confidence interval of 95% you have 2.5% probability in each tail, then you have a value of z = 1.96. This way:

p=50%(as you dont know the proportion you use the maximun error)

You want a precission of e = 1%,

n=z*p (1 - p) / e ^ 2

n= 1.96*(0.5*(1-0.5)/0.01^2) = 4900

Sorry I had a mistake in the formula. The corrected one is:

I.e. for a confidence interval of 95% you have 2.5% probability in each tail, then you have a value of z = 1.96. This way:

p=50%(as you dont know the proportion you use the maximun error)

You want a precission of e = 1%,

n=(z^2)*p (1 - p) / e ^ 2

n= (1.96^2)*(0.5*(1-0.5)/0.01^2) = 9604

In the absence of randomization or probability, you resort to non-probability sampling like snowball, judgmental, convenience, and purposive.

But then you will not be able to make inference from your results.

Dear Hiram 

The best formula is Slovin’s formula ( 1960 ) , but you must know the size of population you study or estimated near by , the formula is :

n= N / ( 1+ N * e^2 )

where

N: population size

n: sample size

e : significance level , for eg. ( 0.05 )

good luck 

One of the best approaches to estimate the needed sample size is Power calculation.

To do power analysis to estimate your sample size, you have to write your hypothesis, and based on that you decide what statistical test you will use. It should be one of the inferential statistics. so you need to determine the following: alpha {standard to be .05}, power [standard to be .80], effect size {small, moderate, or large, each test has its own value, you can find these values in the net}. Then download free programs to calculate the sample size such as G. power.

Dear Khalid Hassen

can you explain how can i cite and reference Slovin's formula? thank you. please if you are voluntary send in my email address   [email protected]. wish you good day.

Dear Desalegn

Slovin's formula widely used to calculate minimum sample size from finite population, but the original reference not available, I suppose you can use other reference as cited by ( ) , I hope the attached file will be useful for this purpose , the auther used the formula and gave Galero-Tejero (2011) as a reference in page 83.

There are good discussion in researchgate about Slovin's formula in the link:

https://www.researchgate.net/post/Who_is_Slovin_and_where_and_how_did_the_Slovins_Formula_for_determining_the_sample_size_for_a_survey_research_originated

Good Luck

Dear RG colleague,

Determining the sample sizes involve resource and statistical issues. Usually, researchers regard 100 participants as the minimum sample size when the population is large. However, In most studies the sample size is determined effectively by two factors: (1) the nature of data analysis proposed and (2) estimated response rate.

For example, if you plan to use a linear regression a sample size of 50+ 8K is required, where K is the number of predictors. Some researchers believes it is desirable to have at least 10 respondents for each item being tested in a factor analysis, Further, up to 300 responses is not unusual for Likert scale development according to other researchers.

Another method of calculating the required sample size is using the Power and Sample size program (www.power-analysis.com).

Regards,

Check Cochran's sample Size formula

https://www.tandfonline.com/doi/full/10.1080/10496491.2017.1346541 check this paper.

Dear Hiram,

You can check this paper where slovin's formula has been used to determine sample size.https://www.emeraldinsight.com/doi/full/10.1108/MRR-06-2017-0173

SLOVEN OR YAMANE'S FORMULA MAY ALSO BE USED

Slovin's and yamanes formulas are the same and it is as follows (N/(1+Ne^2). But the formula is belonging to Slovin as mentioned in Thomas P. Ryan book page 20.

Dear Sir,

Green in (1991) makes two rules of thumb for the minimum acceptable sample size. The first based upon whether you desire to test the overall fit of the regression model and the second based on whether you desire to try out the individual determinants within the model. If you want to examine the model overall, then he recommends a min. sample size of 50+8k, where k = no. of determinants. If you want to examine the individuals predictors then he suggests a min. sample size of 104+k. But when we are interested both in the overall fit and in the contribution of individuals predictors, in such situation Green recommends you to calculate both of the min. sample sizes which i have described above, and use the one that receives the greatest value.

The Sloven formula given has been associated with Taro Yamane

I once read where it was reported that a sample size of 200 is good enough for large population sizes when using inferential statistics