How do i forecast when my dependent and independent variables are numeric and occur on the same time

[Taweesak]

I want to prove if crude oil price are effect to unemployment rate. By the way, the data are time series data, collected every 4 months a year for 10 years. The data are as bellow;

Oil Price	Unemployment
78.01	1.63
82.58	1.61
96.16	1.18
113.96	1.11
87.52	1.65
102.74	1.39
75.23	1.18
53.53	1.33
108.40	2.08
39.10	1.75
85.58	1.17
101.41	0.98
26.00	1.13
97.57	1.32
88.50	0.87
67.01	0.85
99.21	0.83
77.17	0.60
77.69	0.66
74.99	0.63
79.05	0.72
75.36	0.86
70.69	0.58
81.13	0.47
71.00	0.72
84.36	0.74
84.69	0.77
76.47	0.65
84.91	0.89
76.62	1.00
75.03	0.84
78.83	0.61
77.25	0.94
75.73	0.88
80.02	0.92
79.33	0.80
74.27	0.97
81.84	1.08
78.85	0.94
77.27	0.97

so i have to do
1. the stationary test by using ADF test
2. the cointegratio*n test to see the long-term relationship
3. the Grange*r causality test to see if x is grange*r cause y or not
4. the ARIM*A test to predict the future
what syntax should i use?
thank you in advance

[ballardw]

I don't have much experience with the time series procedures but know that most of them will require a date variable to incorporate the "time" portion of the analysis. One date per month is fine but you might want to be consistent about using first of month.