Saturday, September 02, 2006

NACS Demographic Studies

McKibben has done several demographic studies that are available on the net. After looking at several, they had similar if not identical an initial statement on page 2. I was sent the table of three different studies. The titles are:

  • Northwest Allen County Schools: Enrollment Projection January 2002
  • Northwest Allen County Schools January 2004
  • Northwest Allen County Schools: Enrolment Projections July 2006
By demographic convention, a distinction is made between a projection and a forecast; a projection extrapolates the past (and present) into the future with little or no attempt to take into account any factors that may impact this extrapolation (e.g., changes in fertility rates or migration patterns); a forecast results when a projection is modified by judgment to take into account such factors and changes.

The results of this study are to be used as a planning tool, therefore, the ultimate goal is not merely to project the past into the future, but to assess what the likely future may be. Not all of the historical trends have been extended throughout the study period without modification. Forecaster’s judgment has been used to modify some of the demographic trends to more accurately take into account likely changes. Therefore, strictly speaking, this study is a forecast, not a projection; but the two terms will be used interchangeably throughout the report.

When calculating population projections of any type, and particularly for “small” populations such as a school district or its attendance areas, reasonable assumptions must be made as to what the future will bring. The demographic history of the school district in relation to the social and economic history of the area is the starting point and basis of most of these assumptions. The unique nature of each district's demographic composition and rate of change over time must be accounted for and assumed to be factors throughout the life of the projection series. Furthermore, no two populations, particularly at the school district and attendance area level, have exactly the same characteristics.” [1]

The NACS sent just the table and nothing else. From the title on studies for 2002 and 2006, these were projections and McKibben’s written definition of a projection, I assume his studies did not take into account any factors that may impact this extrapolation (e.g., changes in fertility rates or migration patterns). This is contrary to what the superintendent stated at the 10/28 meeting as well as the August 23 meeting.

If this is truly a projection extrapolating the past (and present) into the future, then anyone should be able to do this. Applying the 1989 through 2005 actual enrollment numbers I calculated a nearly perfect correlation of 0.9995. The formula is not linear, but exponential.

The table below lists the actual enrollment reported by NACS to the Indiana Department of Education. I used data from 1989 through 2002 to determine my projection. I assume McKibben's January 2002 study used the same data. Enrollment data for 2006-07 school year is a school count as of August 29, 2006.

Year actual Larsen/accuracy McKibben/accuracy [2]
1989 2,715 2,660 2.1% 2,761 1.7%
1990 2,818 2,789 1.0% 2,840 0.8%
1991 2,856 2,926 -2.4% 2,890 1.2%
1992 3,061 3,068 -0.2% 3,050 -0.4%
1993 3,208 3,218 -0.3% 3,217 0.3%
1994 3,333 3,375 -1.2% 3,159 -5.2%
1995 3,476 3,539 -1.8% 3,478 0.1%
1996 3,759 3,712 1.3% 3,517 -6.4%
1997 3,921 3,893 0.7% 3,912 -0.2%
1998 4,121 4,083 0.9% 4,161 1.0%
1999 4,296 4,282 0.3% 4,314 0.4%
2000 4,474 4,491 -0.4% 4,485 0.2%
2001 4,712 4,710 0.0% 4,795 1.8%
2002 4,961 4,939 0.4% 4,973 0.2%
2003 5,236 5,180 1.1% 5,145 -1.7%
2004 5,519 5,433 1.6% 5,271 -4.5%
2005 5,780 5,698 1.4% 5,364 -7.2%
2006 6,086 5,976 1.8% 5,418 -11.0%

Clearly the 2002
[1] McKibben West Noble Demography Report 2006, page 2

[2] the pdf file was of poor quality. Trying to distinguish between 3, 5, 6, 8 and 9's were difficult. K-5, 6-8 adn 9-12 totals were inputted and totaled and compared to the total on McKibben's projection. There were differences and are due to poor legibility of all values.

0 Comments:

Post a Comment

Links to this post:

Create a Link

<< Home



NBC-33 Debate poll results from 2002