Interpreting weighting scheme results in modern reports – Alida Help Center

Learn about how to interpret the Weight Scheme Results report, and determine the effectiveness of your weighting scheme.

Number	Option	Description
1	Combination of variables	The number of variables that are combined and analyzed by the weighting scheme. For example, you add the following fields to a weighting scheme: Gender (Male, Female) Age (18-34, 34-54, 55+) The combined value is 6 because each combination of the two fields must be weighted: Male 18-34 Male 34-54 Male 55+ Female 18-34 Female 34-54 Female 55+ The maximum value for Combination of Variables is 250, 000.
2	Last calculated	The date and time when the weighting scheme was last calculated. This value is set when the weighting scheme is created and updated whenever the weighting scheme is edited.
3	Target percentage achieved	Indicates that the weighting calculations were successful and all answers in the weighting scheme could be adjusted to their target percentages.
4	Weighting efficiency	Weighting efficiency is a percentage value ranging from 0% to 100% that reflects the extent to which weights have impacted the variability of the survey data, and how close the weighted sample is to the target population. A 100% value means that weighting only made minor adjustments to the sample, and indicates that the sample was probably already close to representing the target population. A lower value (60% or 70%) means that weighting made substantial adjustments to correct the sample. This typically occurs when some responses are given very high or very low weights to adjust for sample discrepancies. If the value is lower than 70%, you should reevaluate your weighting scheme because it indicates that extreme weights have been applied to responses and you are likely introducing biases. The weighting efficiency affects the number or responses analyzed in your report. If the weighting efficiency is 80%, this means that weighting reduces the sample by 20%. If the sample contained 1000 responses, it is as if you have removed 200 responses from the sample (1000 * 80% = 800). This may be an important factor to consider when there is cost associated with acquiring each respondent.
5	Weighted base	The total number of responses after weighting is applied. This value represents the total weight applied to the sample.
6	Unweighted base	The total number of responses before weighting is applied.
7	Effective base	The adjusted sample size after applying weights. This value accounts for variability in weights (i.e., how weights differ across respondents). It provides a measure of how representative or reliable the weighted data are. The closer the effective base is to the unweighted base, the less the weighting scheme is manipulating the sample data. An effective base value that is similar to the unweighted base value indicates that the weighted sample closely resembles the unweighted sample. This means that the weighted sample retains the precision of the unweighted sample. If the effective base value is much smaller than the unweighted sample, this is an indication that some responses have been given much more influence than others. The effect of this is that the weighted sample loses some of the precision found in the unweighted sample.
8	Maximum weight	The maximum weight factor applied to a variable or variable combination. The recommended maximum weight is less than 3. If you specified a Maximum Weight Factor Limit for your weighting scheme, the Maximum weight will not exceed this value.
9	Minimum weight	The minimum weight factor applied to a variable or variable combination. The recommended minimum weight factor is greater than 0.5. If you specified a Minimum Weight Factor Limit for your weighting scheme, the Minimum weight will not be less than this value.

Related articles