One datapoint in time … or two … or more
When there is just one datapoint it is possible to know something … but not really very much. As soon as there is a second datapoint there is another dimension of analysis and understanding that is possible. There is now an ability to have some comparative direction … is the item more or less … what is the change … what is the trend.
TVM uses time series of key items to gain an understanding of what is happening in the community.
Multiple datapoints provides a time series
A time series may be build from multiple datapoints over time. Raw data plotted as a time series tells lots of stories
When I made a very simple plot of prices of shrimp in the New York market month by month over a period of nearly thirty years .. from 1946 to 1974 ... I gained a perspective of the shrimp industry better than most. I believe this enabled me to interpret the history of the industry correctly, and because of this I was able to predict how the oil shock turmoil of the 1970s would impact our company ... a major producer of marketer of shrimp worldwide ... and how we should position ourselves for success.
Time series data … trends are important
Time series trends are great indicators of progress ... or not. Time series are simple, clear and powerful. While it is possible to do advanced statistical manipulation ... simple and clear time series tables and charts work very powerfully as well ... maybe better.
Time series are also very easy to manipulate so that the result is serious misinformation. Objective rigorous analysis will produce good results … but the analysis may be rigged to produce data that suggests completely wrong trends.
There are ways to prevent time series data being used to suggest one trend when the opposite is true. One technique is to included both the short term detail together with the long term trend on a longer time scale.
Multiple baseline time series
A plot of a single parameter shows how this parameter has changed over time ... but in isolation does not show what might have been the cause of any changes. Plotting multiple variable may show something about cause and effect. While this may be done by simple visualization for a couple of variables, a more rigorous mathematical approach is needed for large scale multivariate analysis.
In any specific situation is is possible to plot different datasets as a time series, and “see” how each element changes relative to the other. The goal is not to have an academically rigorous statistical correlation, but to find out what seems to relate to what.
TVM uses time series analysis … and TVM uses the datasets that may be related to very specific activities or locations. An excessive amount of external variables makes for too much “noise” in the data and make the comparative plots meaningless.
Changes over time
In the section on spatial attributes of data, the following was said:
In reports about the lives of people in poor rural communities in developing countries, the distance women and girls have to walk to collect water and to collect firewood is usually a big item. This is actually very important … and is about the spatial situation in the community. Distance to water and distance to firewood are important facts.
Something else is important … it is the matter of time and the question of change. Twenty years ago … or thirty years ago … the water and firewood was only a short distance away. Over time there have been changes, and now the situation is dire. But this change did not happen suddenly, this change happened over a period of many years … and nobody took any notice. The fact that nobody took any notice is an indication of failure of management and metrics.
Some important time series
Market prices are a leading indicator of market conditions and other broader issues in the community. In economies that are reliant on subsistence farming, a combination of high food prices and low livestock prices is a reliable indicator of emerging famine conditions. People in remote villages know this … and have known this for thousands of years … while development experts and government decision makers may not yet have grasped the full significance of these indicators.
Data identified the problem … what to do?
Tracking market prices helps to identify issues in the economic performance … but what to do. Food price increases can result in malnutrition, starvation and death … and the “invisible hand” may in time respond with more supply of food, but only if there is buying power and market demand as well as high prices. Markets about money profit do not have any way to monetize the value of survival and saving lives … but a mechanism is needed in situations where basic needs for survival cannot be satisfied. Pre-positioned security stocks of food are a part of the answer … but there also needs to be other elements of organization, management and metrics in place to provide for oversight and accountability.