Because the ACS is a survey, data reported are estimates, not exact counts. Some tables included in each release provide information about the underlying details. This data is mostly used by the Census Bureau to monitor the accuracy and reliability of the ACS and to inform updates to the survey design and methodology. We include them in Census Reporter because it's easier to keep them when we process the other data than to skip over them. If you're using them for anything, let us know.
Tables beginning with 98 organize various data which survey methodology experts can use to assess the quality of data collected by the ACS. Data in these tables is limited to large summary levels: the nation, individual states, and occasionally counties and/or places. Most users of Census Reporter will not have any need to use these tables.
Imputation is the process of correcting for "non-response": that is, when answers on the ACS questionnaire are not completed, or are completed with invalid values. Allocation is one strategy for responding to the non-response, by using statistical procedures to impute ("fill in") missing values.
Tables beginning with 99 provide the allocation rate for nearly every question on the ACS questionnaire. Values are available for nearly all summary levels.