Importance of Weighting in Secondary Data
According to Pike (2008), the weighting is often done in secondary research studies that involve multiple values of assessment. Weighting allows us to understand the importance of different facets of analysis. In public health research, the weighting is instrumental in aligning a study’s findings to what researchers already know about a population. In this regard, weighting increases the relevance of research data. Furthermore, it helps to eliminate distortions in secondary data analysis, thereby allowing enough room for researchers to focus on what is important (aspects that have the most weight). Howell (2012) also says that it helps to communicate the precision of research data.
How to use Weighting for Secondary Data
Consider the following data set, which provides information about the types of insurance cover Americans have.
|Respondent||Type of Insurance|
In the above sample, we see that 80% of the respondents have employer insurance. Since the point of conducting health, assessments is to draw conclusions about entire populations and not just samples (Shi & Johnson, 2014), we would use the weighting technique to say employer insurance is the most common type of insurance as opposed to saying 80% of Americans use employer insurance.
In our second example, the following data explain the employment status of the respondents.
|Respondent||Type of Insurance||Employment Status|
A quick assessment of the above data shows that the data set is unrepresentative of all Americans because all the people who had employer insurance were employed and those who did not have this type of insurance were unemployed. We would use the weighted method to assign a weight to each group of the respondent. Hypothetically, if 60% of Americans have a job and their representation in our sample is 80%, we would say that the weight of employer insurance is 60%/80% = 0.75. Conversely, the weight for the unemployed is 40%/20% = 2. Thus, our overall data is
|Respondent||Type of Insurance||Weight|
The rationale for Weighting in the Above Examples
In the first example, it was important to weigh the data because health research is often aimed at drawing conclusions about the general population and not just the sample size. Therefore, the rationale for weighting the data was to generalize it (Whitener, Van Horne, & Gauthier, 2005).
In the second example, it was important to weigh the data because we needed to find out how much the sample population represented the American public.
Additional Insight and Validation of Idea
Although weighting helps to fix many problems associated with secondary data, it is important to understand that it does not fix all problems. For example, it would not fix any biases associated with age group or income group discrepancies of the sampled population (Pike, 2008). Howell (2012) supports this fact by saying that weighting would not help us to understand other problems associated with the sampled data. Therefore, it would only fix those problems that researchers believe that if solved, would lead to the creation of representative data.
Howell, D. C. (2012). Treatment of missing data—Part 1. Web.
Pike, G. R. (2008). Using weighting adjustments to compensate for survey nonresponse. Research in Higher Education, 49(2), 153–171. Web.
Shi, L., & Johnson, J. A. (Eds.). (2014). Novick & Morrow’s public health administration: Principles for population-based management (3rd ed.). Burlington, MA: Jones & Bartlett.
Whitener, B. L., Van Horne, V. V., & Gauthier, A. K. (2005). Health services research tools for public health professionals. American Journal of Public Health, 95(2), 204–207. Web.