Home
/
Math
/
b) It i s not reliab le to u se the line best fi t to e stimat e the h eigh t of a 33 yea r old. G ive one reaso n why. Heights of students of different ages

Question

b) It i s not reliab le to u se the line
best fi t to e stimat e the h eigh t of a
33 yea r old. G ive one reaso n why.
Heights of students of different ages

b) It i s not reliab le to u se the line best fi t to e stimat e the h eigh t of a 33 yea r old. G ive one reaso n why. Heights of students of different ages

expert verifiedVerification of experts

Answer

4.0367 Voting
avatar
CoreyMaster · Tutor for 5 years

Answer

One fundamental reason would be, using linear regression to predict heights of individuals beyond the initially provided data range (extrapolation) could produce unreliable results, because there are chances that actual age-height birfurcation may stagger outside the known "age and height" zones coupled with the certainty that height's correlation with age would highly likely plateau upon reaching adulthood.

Explain

## Step 1: Analyze the Problem<br /><br />The question involves the topic of linear regression. It pertains to using a best fit line, which represents past data, to estimate the height of a 33-year-old. This assumption is based on the information derived from the height and age of students which can vary extensively.<br /><br />## Step 2: Identify the Pitfalls <br /><br />Having known the variation that can occur in these variables (age and height), using historical values to predict/estimate future outcomes isn't a fully reliable method, for a few reasons.<br /><br />## Step 3: Reasoning<br /> <br />One primary factor stands out when using linear regression on variables such as height and age that, past a certain stage, these variables don't follow a linear path. For instance, when a child is growing, there might be a steady increase in height with age, but after reaching full adult height, this relationship tends to plateau, forming more of a horizontal line. Hence, linear regression no longer is a good choice.<br /><br />Another important reason could be 'extrapolation'. Extrapolation involves prediction beyond the zone honoured by data. In other words, when making a prediction out of the range of given age and heights, linear regression can produce misleading results, because the correlation pattern may significantly change beyond the known range supplied in the dateset.
Click to rate: