Most of my variable are insignificant, what can i do? hurry

[Jenny0917]

hello, i am doing a estimation about what fator effect on chinese international routes of civil aviation industry.
dependent variable:International routes
explanatory varibales:
UPt - Urban population in year t, (ten thousands)
GDPt - GDP in year t, (hundred million yuan)
NDRt - Number of domestic routes in year t
AIRPt - Number of airport in year t
AIRCt - Number of aircraft in year t
INt - Number of inbound tourist in year t, (ten thousands)
OUTt - Number of outbound tourist in year t, (ten thousands)
FIt - Foreign investment inflows in year t, (then thousands US dollars)
Dummy – value equal to 1 since year 2002 to 2013, value equal to 0 from 1992 to 2001.
µt - an error term.

1, i checked : they are not stationary then i transform all the variable in logarithms. Is it correct ?

2, only 4 and 8 are significant. then should i check the cointegration or multilinearity test or Heteroskedasticity test ?
Dependent Variable: LOG(INROUT)

Method: Least Squares
Sample (adjusted): 1994 2013
Included observations: 20 after adjustments

Variable Coefficient Std. Error t-Statistic Prob.

C 12.31529 7.982445 1.542797 0.1511
LOG(_1_UP) -1.963980 1.126591 -1.743294 0.1091
LOG(_3_NDR) 0.381719 0.238948 1.597497 0.1385
LOG(_4_AIRP) -1.580454 0.555987 -2.842610 0.0160
LOG(_5_AIRC) 0.492010 0.296727 1.658124 0.1255
LOG(_6_IN) 0.648699 0.388146 1.671276 0.1228
LOG(_7_OUT) 0.381913 0.241876 1.578962 0.1426
LOG(_8_FI) 0.448041 0.185696 2.412772 0.0344
DUMMY 0.031579 0.107163 0.294678 0.7737

R-squared 0.991027 Mean dependent var 5.264650
Adjusted R-squared 0.984501 S.D. dependent var 0.533269
S.E. of regression 0.066389 Akaike info criterion -2.284396
Sum squared resid 0.048483 Schwarz criterion -1.836316
Log likelihood 31.84396 Hannan-Quinn criter. -2.196926
F-statistic 151.8604 Durbin-Watson stat 2.818355
Prob(F-statistic) 0.000000

3, But when i do the Johansan cointegration test (include dummy vriable) it shows that " Insufficient number of observation".

4 What can i do to make the model signifcant or what test shoud i use?

I am really in a hurry, someone can help me ? Thank you so much !