Data fusion represents the matching of two or more databases at the respondent level to create a unified database. The matching uses information common to both surveys, called “linking variables” or “fusion hooks,” to pair up respondents from the respective databases. The end product is a respondent-level data set containing information from both sets of data, simulating single-source data.
By cross referencing results from two or more databases, data fusion delivers insights that are unavailable from the individual databases. Data fusion is cost effective, as it requires no additional primary research, fits into existing analysis systems, and enhances existing data. The fused data are actionable, giving clients the ability to link syndicated and custom data with currency data such as the US National People Meter TV panel.
Nielsen conducts extensive validation of its data fusion products, and this work suggests that the results produced are very reliable, allowing clients to work confidently with fused data.
