May 9, 2009 7:17 PM
As compared to the shift from manual work to automation, monthly analysis at UZRE.uz has undergone an unnoticeable change - a shift from aggregate average to average of averages. Such treatment of figures came more favorable under varying sample sizes.
Since the major focus of the periodical is estimating the average price deviation, our analysis has insatiable hunger for as many available apartments for sale as possible and available data is subject to changes in terms of sample sizes, etc.
While, UZRE.uz analyzes Tashkent by three zones, we have two ways of estimating average at our discretion: Aggregate average (used up until April), which we derive by dividing the total sum by total sample size, and average of averages (newly adopted).
Aggregate average may yet show an increase even when the average of each zone declines. It is not quite pleasing to the eye, when you notice a decline in each zone, but the overall performance indicates an increase, the reason being varying sample size.
In improving our estimates, average of averages contributes greatly. It is quite straightforward, and its simple arithmetic does not allow over or underestimate the average.
Since, monthly report presents cumulative data, a change was taken reciprocally, and is represented in April report.
| # rooms | Price | Change |
|---|---|---|
| 1-room | $ 19,037 | -3.78 % |
| 2-room | $ 27,299 | -4.45 % |
| 3-room | $ 39,169 | -4.53 % |
| 4-room | $ 46,490 | -1.96 % |