8
estimated coefficients and their significance are found in Table 5. Note lack of major
significance for total bathrooms which would be explained by the higher correlation with
bedrooms and square footage. The months of May, June, and July are not noted as significant
in this model as well as the Covid variable. Covid would be explained by high correlation with
the year 2020 variable. Insignificant variables were included in the model as they do contribute
to a higher R
2
value, meaning they do help explain the variance in the model. All other variables
are significant with 99% confidence.
The estimated model indicates that for every percent increase of price, time on market
will increase by 0.321%. For every year older that a house is, there will be a decrease of 0.1% of
time on market. For every bedroom included in a property, time on market decreases by 7%.
For every 100 square feet, DOM increases 1.4%. Every year after 2010 decreased time on
market compared to 2010. DOM in 2011 decrease by 56.3% in comparison to 2010; similarly,
2012 decreased by 76.3%, 2013 by 78.2%, 2014 by 82.8%, 2015 by 122.0%, 2016 by 164.1%,
2017 by 183.7%, 2018 by 177.3%, 2019 by 172.8%, and 2020 decreased DOM by 196.3%. All
months that held significance increased time on market as compared to April. Compared to
April, for example, January increased DOM by 40.7%, February by 20.2%, March by 14.1%,
August by 22.3%, September by 18.8%, October by 30.9%, November by 36% and December by
34.9%. The positive premium quartiles all decreased time on market compared to those houses
that sold at asking price or less. Quartile 1 or Premium1 decreases DOM by 68.7%, Premium2
by 81.3%, Premium3 by 82.7%, and Premium4 by 71.6%. The InvAve variable indicates an
increase of DOM of 12.4% when compared to those houses that sold below annual average