Analysis of Data for Measuring Food Availability, Access and Nutritional Status (Topic 8)

Training Material
Subtitle: 
[Data Verification: Graphical Approaches]

This topic aims to refresh ideas on the rationale behind the graphical presentation of data using tools, like bar diagrams, histogram and box plots. An important use of such graphical tools is to study the nature of distributions, identify outliers and judge suitability for statistical analysis, like ordinary least squares. Though most economic variables follow positively skewed distributions, and hence mean-based estimates of averages have limited relevance in studies on average levels of living and food security, empirical studies quite often use mean based estimates to assess the magnitude of the problem and make policy recommendations. Therefore, it is very important to study the nature of distribution of a variable in a sample before deciding on a measure of average. The topic considers distinctions between a bar diagram and a histogram, the rationale involved in drawing a histogram for a frequency distribution with unequal class intervals, and steps in drawing a scatter plot and plotting bivariate relations like Engel functions.

The main message of this topic is the need for would be have an integrated perspective and to verify univariate distributions and bivariate relations separately before deciding on functional forms of relationships between variables in FNS analysis.
 

Download documentSize
Analysis of Data for Measuring Food Availability, Access and Nutritio...536.46 KB
Analysis of Data for Measuring Food Availability, Access and Nutritio...315.07 KB
Date and language
Jan 2012
English
Personal Author
Other information
Form: 
Slides
Form: 
Web resource

Share this