calculation of sampling errors mics3 data analysis and report writing workshop
TRANSCRIPT
Calculation of Sampling Errors
MICS3 Data Analysis and Report Writing Workshop
Background
The sample selected in a survey is one of the many samples that could have been selected (with same design and size).
Sampling errors are measures of the variability between all possible samples, which can be estimated from survey results.
Background
Calculation of sampling errors is very important;
-Provides information on the reliability of your results
-Tells you the ranges within which your estimates most possibly fall
-Provide clues as to the sample sizes (and designs) to be selected in forthcoming surveys
Background
MICS3 sample designs are complex designs, usually based on stratified, multi-stage, cluster samples.
It is not possible to use straightforward formulae for the calculation of sampling errors. Sophisticated approaches have to be used
New versions of SPSS (13 or 14) is used for this purpose.
SPSS uses Taylor linearization method of variance estimation for survey estimates that are means or proportions.
This approach is used by most other package programs: Weswar, Sudaan, Systat, EpiInfo, SAS
Background
In MICS3, the objective is to calculate sampling errors for a selection of variables, for the national sample, as well as selected sub-populations, such as urban and rural areas, and regions
Sampling errors will be presented as part of the final report, in an appendix
BackgroundSampling ErrorsStandard errors, coefficients of variation, design effects (DEFF), square root of design effects (DEFT) and confidence intervals for selected variables, Country, Year
Table Value (p)Standard error (SE)
Coefficient of variation
Design effect (DEFF)
Square root of design effect
(DEFT)Unweighted
count p - 2SE p + 2SE
Household availability of insecticide treated nets CH.10 0.000 0.000Iodized salt consumption NU.5 0.000 0.000
Improved source of drinking water EN.1 0.000 0.000Sanitary means of excreta disposal EN.5 0.000 0.000Net primary school attendance rate ED.3 0.000 0.000Net secondary school attendance rate ED.4 0.000 0.000
Skilled attendant at delivery RH.5 0.000 0.000Antenatal care RH.3 0.000 0.000
Underweight prevalence NU.1 0.000 0.000Had acute respitory infection CH.6 0.000 0.000Antibiotic treatment of suspected pneumonia CH.7 0.000 0.000Diarrhoea in last two weeks CH.4 0.000 0.000Received ORT or increased fluids and continued feeding CH.5 0.000 0.000Children sleeping under ITNs CH.11 0.000 0.000Fever in last two weeks CH.12 0.000 0.000Antimalarial treatment CH.12 0.000 0.000Support for learning CD.1 0.000 0.000Birth registration CP.1 0.000 0.000
UNDER-5s
Confidence limits
HOUSEHOLDS
HOUSEHOLD MEMBERS
WOMEN
BackgroundSampling ErrorsStandard errors, coefficients of variation, design effects (DEFF), square root of design effects (DEFT) and confidence intervals for selected variables, Country, Year
Table Value (p)Standard error (SE)
Coefficient of variation
Design effect (DEFF)
Square root of design effect
(DEFT)Unweighted
count p - 2SE p + 2SE
Household availability of insecticide treated nets CH.10 0.000 0.000Iodized salt consumption NU.5 0.000 0.000
Improved source of drinking water EN.1 0.000 0.000Sanitary means of excreta disposal EN.5 0.000 0.000Net primary school attendance rate ED.3 0.000 0.000Net secondary school attendance rate ED.4 0.000 0.000
Skilled attendant at delivery RH.5 0.000 0.000Antenatal care RH.3 0.000 0.000
Underweight prevalence NU.1 0.000 0.000Had acute respitory infection CH.6 0.000 0.000Antibiotic treatment of suspected pneumonia CH.7 0.000 0.000Diarrhoea in last two weeks CH.4 0.000 0.000Received ORT or increased fluids and continued feeding CH.5 0.000 0.000Children sleeping under ITNs CH.11 0.000 0.000Fever in last two weeks CH.12 0.000 0.000Antimalarial treatment CH.12 0.000 0.000Support for learning CD.1 0.000 0.000Birth registration CP.1 0.000 0.000
UNDER-5s
Confidence limits
HOUSEHOLDS
HOUSEHOLD MEMBERS
WOMEN
Value of the estimate should be the same as that in the corresponding table
BackgroundSampling ErrorsStandard errors, coefficients of variation, design effects (DEFF), square root of design effects (DEFT) and confidence intervals for selected variables, Country, Year
Table Value (p)Standard error (SE)
Coefficient of variation
Design effect (DEFF)
Square root of design effect
(DEFT)Unweighted
count p - 2SE p + 2SE
Household availability of insecticide treated nets CH.10 0.000 0.000Iodized salt consumption NU.5 0.000 0.000
Improved source of drinking water EN.1 0.000 0.000Sanitary means of excreta disposal EN.5 0.000 0.000Net primary school attendance rate ED.3 0.000 0.000Net secondary school attendance rate ED.4 0.000 0.000
Skilled attendant at delivery RH.5 0.000 0.000Antenatal care RH.3 0.000 0.000
Underweight prevalence NU.1 0.000 0.000Had acute respitory infection CH.6 0.000 0.000Antibiotic treatment of suspected pneumonia CH.7 0.000 0.000Diarrhoea in last two weeks CH.4 0.000 0.000Received ORT or increased fluids and continued feeding CH.5 0.000 0.000Children sleeping under ITNs CH.11 0.000 0.000Fever in last two weeks CH.12 0.000 0.000Antimalarial treatment CH.12 0.000 0.000Support for learning CD.1 0.000 0.000Birth registration CP.1 0.000 0.000
UNDER-5s
Confidence limits
HOUSEHOLDS
HOUSEHOLD MEMBERS
WOMEN
Standard error is the square root of the variance – a measure of the variability between all possible samples
BackgroundSampling ErrorsStandard errors, coefficients of variation, design effects (DEFF), square root of design effects (DEFT) and confidence intervals for selected variables, Country, Year
Table Value (p)Standard error (SE)
Coefficient of variation
Design effect (DEFF)
Square root of design effect
(DEFT)Unweighted
count p - 2SE p + 2SE
Household availability of insecticide treated nets CH.10 0.000 0.000Iodized salt consumption NU.5 0.000 0.000
Improved source of drinking water EN.1 0.000 0.000Sanitary means of excreta disposal EN.5 0.000 0.000Net primary school attendance rate ED.3 0.000 0.000Net secondary school attendance rate ED.4 0.000 0.000
Skilled attendant at delivery RH.5 0.000 0.000Antenatal care RH.3 0.000 0.000
Underweight prevalence NU.1 0.000 0.000Had acute respitory infection CH.6 0.000 0.000Antibiotic treatment of suspected pneumonia CH.7 0.000 0.000Diarrhoea in last two weeks CH.4 0.000 0.000Received ORT or increased fluids and continued feeding CH.5 0.000 0.000Children sleeping under ITNs CH.11 0.000 0.000Fever in last two weeks CH.12 0.000 0.000Antimalarial treatment CH.12 0.000 0.000Support for learning CD.1 0.000 0.000Birth registration CP.1 0.000 0.000
UNDER-5s
Confidence limits
HOUSEHOLDS
HOUSEHOLD MEMBERS
WOMEN
Coefficient of variation (relative error) is the ratio of SE to the estimate
BackgroundSampling ErrorsStandard errors, coefficients of variation, design effects (DEFF), square root of design effects (DEFT) and confidence intervals for selected variables, Country, Year
Table Value (p)Standard error (SE)
Coefficient of variation
Design effect (DEFF)
Square root of design effect
(DEFT)Unweighted
count p - 2SE p + 2SE
Household availability of insecticide treated nets CH.10 0.000 0.000Iodized salt consumption NU.5 0.000 0.000
Improved source of drinking water EN.1 0.000 0.000Sanitary means of excreta disposal EN.5 0.000 0.000Net primary school attendance rate ED.3 0.000 0.000Net secondary school attendance rate ED.4 0.000 0.000
Skilled attendant at delivery RH.5 0.000 0.000Antenatal care RH.3 0.000 0.000
Underweight prevalence NU.1 0.000 0.000Had acute respitory infection CH.6 0.000 0.000Antibiotic treatment of suspected pneumonia CH.7 0.000 0.000Diarrhoea in last two weeks CH.4 0.000 0.000Received ORT or increased fluids and continued feeding CH.5 0.000 0.000Children sleeping under ITNs CH.11 0.000 0.000Fever in last two weeks CH.12 0.000 0.000Antimalarial treatment CH.12 0.000 0.000Support for learning CD.1 0.000 0.000Birth registration CP.1 0.000 0.000
UNDER-5s
Confidence limits
HOUSEHOLDS
HOUSEHOLD MEMBERS
WOMEN
Design effect is the ratio between the SE using the current design and the SE that would result if a simple random sample was used. A DEFT value of 1.0 indicates that the sample is as efficient as a SRS
BackgroundSampling ErrorsStandard errors, coefficients of variation, design effects (DEFF), square root of design effects (DEFT) and confidence intervals for selected variables, Country, Year
Table Value (p)Standard error (SE)
Coefficient of variation
Design effect (DEFF)
Square root of design effect
(DEFT)Unweighted
count p - 2SE p + 2SE
Household availability of insecticide treated nets CH.10 0.000 0.000Iodized salt consumption NU.5 0.000 0.000
Improved source of drinking water EN.1 0.000 0.000Sanitary means of excreta disposal EN.5 0.000 0.000Net primary school attendance rate ED.3 0.000 0.000Net secondary school attendance rate ED.4 0.000 0.000
Skilled attendant at delivery RH.5 0.000 0.000Antenatal care RH.3 0.000 0.000
Underweight prevalence NU.1 0.000 0.000Had acute respitory infection CH.6 0.000 0.000Antibiotic treatment of suspected pneumonia CH.7 0.000 0.000Diarrhoea in last two weeks CH.4 0.000 0.000Received ORT or increased fluids and continued feeding CH.5 0.000 0.000Children sleeping under ITNs CH.11 0.000 0.000Fever in last two weeks CH.12 0.000 0.000Antimalarial treatment CH.12 0.000 0.000Support for learning CD.1 0.000 0.000Birth registration CP.1 0.000 0.000
UNDER-5s
Confidence limits
HOUSEHOLDS
HOUSEHOLD MEMBERS
WOMEN
Upper and lower confidence limits are calculated as p +/- 2.SE
Indicate the ranges within which the estimate would fall in 95 percent of all possible samples of identical design and size
How SPSS works
COMPLEX SAMPLES module
Can be used to select a sample, or indicate the design of the sample from which the data set comes, so that sampling error estimates can be calculated
Calculations can be done for means and proportions, ratios, frequencies and crosstabs. Also possible to use general linear models and logistic regression.
How SPSS works
Prepare an analysis file to indicate the parameters that define the sample design.CSPLAN ANALYSIS /PLAN FILE='micsplan.csplan' /PLANVARS ANALYSISWEIGHT=hhweight /PRINT PLAN /DESIGN STRATA= strat CLUSTER= HH1 /ESTIMATOR TYPE=WR.
Using the plan file, calculate sampling errors.Complex Samples Descriptives.CSDESCRIPTIVES /PLAN FILE = 'micsplan.csplan' /SUMMARY VARIABLES =treated iodized /MEAN /STATISTICS SE CV COUNT DEFF DEFFSQRT /MISSING SCOPE = ANALYSIS CLASSMISSING = EXCLUDE.
Problems with using SPSS
Need to pair clusters and create pseudo-strata.
Cannot handle normalized weights – multiply the weights by 1,000,000 before analysis.
Provides estimates for subpopulations only when the data file used contains only cases for the subpopulation in question
Provides incorrect confidence limits
Cannot report on sampling errors for variables coming from different data sets
SPSS OutputUnivariate Statistics
.0285 .00608 .213 6.418 2.533 4811
.4722 .01050 .022 2.119 1.456 4786
Household availability ofinsecticide treated nets
Iodized salt consumption
MeanEstimate
StandardError
Coefficientof Variation Design Effect
Square RootDesign Effect
UnweightedCount
Univariate Statistics
.6920 .02211 .032 67.767 8.232 29560
.9306 .00510 .005 11.919 3.452 29560
.6212 .00658 .011 .828 .910 4504
.8157 .00788 .010 2.240 1.497 5422
Improved source ofdrinking water
Sanitary means ofexcreta disposal *
Net primary schoolattendance rate
Net secondary schoolattendance rate
MeanEstimate
StandardError
Coefficientof Variation Design Effect
Square RootDesign Effect
UnweightedCount
Univariate Statistics
.8309 .01582 .019 2.151 1.467 1209
.7836 .01691 .022 2.036 1.427 1209
Skilled attendantat delivery
Antenatal care
MeanEstimate
StandardError
Coefficientof Variation Design Effect
Square RootDesign Effect
UnweightedCount
SPSS OutputUnivariate Statistics
.1699 .00786 .046 1.312 1.146 2995
.0132 .00199 .150 .956 .978 3167
.3507 .05078 .145 .487 .698 44
.1340 .00760 .057 1.576 1.255 3167
.3311 .02156 .065 .904 .951 432
.0097 .00404 .416 5.375 2.318 3167
.0743 .00630 .085 1.824 1.350 3167
.0168 .00748 .446 .810 .900 240
.6178 .01103 .018 1.630 1.277 3167
.8688 .00872 .010 2.114 1.454 3167
Underweight prevalence
Had acute respitoryinfection
Antibiotic treatment ofsuspected pneumonia
Diarrhoea in last twoweeks
Received ORT orincreased fluids ANDcontinued feeding
Children sleeping underITNs
Fever in the last twoweeks
Antimalarial treatment
Support for learning
Birth registration
MeanEstimate
StandardError
Coefficientof Variation Design Effect
Square RootDesign Effect
UnweightedCount