Application of Data Smoothing Techniques to Trend Analysis of Large-Scale Survey Data

One of the problems in identifying trends in time series or spatial data is that there are usually so many irregularities or random fluctuations that the underlying trend is difficult to discern. Smoothing techniques can be used to reduce local irregularities (local in the sense of being close in time or geography) so that the underlying trend becomes clear. Although these techniques are widely used in the physical sciences, they are seldom used in the social sciences or in spatial (GIS) applications. During this internship, the intern will examine a range of different smoothing techniques to examine trends in continuous monitoring data collected through Canada’s largest on-line school survey called Tell Them From Me (TTFM). This work will lead to further research in the application of these techniques to smoothing spatial data. Tell Them From Me is the product of the industry partner, the Learning Bar, which creates monitoring and assessment tools for schools.

Faculty Supervisor:

Dr. J. Douglas Willms

Student:

Richard Chan

Partner:

Learning Bar Inc.

Discipline:

Engineering

Sector:

Information and communications technologies

University:

University of New Brunswick