While shielding will help, there are still many possibilities for noise currents. But since you are running a laptop that also maybe core to the problem. The a3967 minimum input high spec is 3.5V at 5V supply voltage. That would show up when sending a true set of step and direction signals. But that board alone, doesn't support a requirement allegro notes on the A3967 and I explain below.
Motor jitter when the cable is connected, more likely is false triggers in the chopper circuit. That maybe because of attaching a long antenna to the gnd of the board to the pc lpt port, or improper gnd wiring, or gnd wireing layout. Lots of possibilities.
Make sure that your power and gnd wiring is short and direct, doesn't physically lay next to motor wires. And that you have wired the PC LPT port pins properly to the board.
if you look at the schematics posted on my website, you will see my designs use a input circuit to condition the driver chip. One example
http://pminmo.com/ss3977/ss3977schematic.pdf They use of the schmitt trigger chip (74HC14) is more friendly to PC's (if necessary the user can use a 74HCT14 for a laptop) plus an RC filter for the incomming step-dir-enable signals. In addition, the power gnd path to the power supply is on a different connector than the logic interface to the pc. There should be almost no gnd current to the PC parallel port, a few millamps at most. The input circuit arrangement also provides a level of isolation (4k to 10k) if the driver chip would fail and internally short the motor supply to the input pins.
That circuit also meets the requirement that Allegro notes on many of their chips, including the A3967 of the need for a low impedance source near the chip. The LPT driver feet away from the PC doesn't.