You are estimating a model explaining out-of-pocket health care spending as a function of age, number of health conditions, female dummy, and years of education. Note that coefficient for age is not significant with low t-score.
. reg health_care_cost age num_hconditions female raedyrs if age<=80
Source | SS df MS Number of obs = 552
-------------+---------------------------------- F(4, 547) = 2.84
Model | 859820811 4 214955203 Prob > F = 0.0238
Residual | 4.1427e+10 547 75734668.9 R-squared = 0.0203
-------------+---------------------------------- Adj R-squared = 0.0132
Total | 4.2287e+10 551 76745344.3 Root MSE = 8702.6
---------------------------------------------------------------------------------
health_care_c~t | Coef. Std. Err. t P>|t| [95% Conf. Interval]
----------------+----------------------------------------------------------------
age | -123.1563 219.3149 -0.56 0.575 -553.9588 307.6462
num_hconditions | 887.5861 330.8229 2.68 0.008 237.7472 1537.425
female | 1698.277 808.0694 2.10 0.036 110.9782 3285.576
raedyrs | 185.5531 108.8725 1.70 0.089 -28.30624 399.4124
_cons | 8668.253 17063.78 0.51 0.612 -24850.3 42186.8
---------------------------------------------------------------------------------
Remove age since it is an irrelevant variable. |
||
Age is neither omitted not irrelevant variable |
||
Keep age since it is theoretically important variable and removing age would lead to omitted variable bias. |
||
Age is both omitted and irrelevant variable |
You are estimating a model explaining out-of-pocket health care spending as a function of age, number of health conditions, female dummy, and years of education.
Note that coefficient for age is not significant, with a low t-score.
---
Correct answer:
Keep age since it is theoretically important variable and removing age would lead to omitted variable bias.
---
In the specified model, age is a very important variable that may explain out-of-pocket health care expenditure. The results suggest that the coefficient is insignificant, and it has a low t-score.
However, this doesn't mean that the variable should be dropped. It can't be removed from the model, as it is theoretically relevant. Though statistically it has been proved to be insignificant, dropping it would lead to omitted variable bias.
Get Answers For Free
Most questions answered within 1 hours.