In open circuit test, magnetizing circuit is highly inductive with rated voltage, so the pf is low.  

In SC test effect of magnetizing shunt branch is negligible, as full load current is circulated at very much reduced voltage. Hence pf will be high. Refer to standard books on transformers.

I agree with Krishnarayalu.

To further explain the situation, under open circuit condition, if you refer to the primary referred equivalent circuit of the transformer, then due to absence of load, the circuit is completed through the primary resistance & primary leakage impedance in series with the parallel combination of magnetizing reactance and core loss resistance. Since the core loss is not large, the core loss resistance is high. Thus, the magnetizing reactance dominates and the circuit current is highly lagging. Hence poor pf.

Under short circuit condition, the secondary is shorted. hence the circuit is completed through the primary resistance, primary leakage reactance, secondary resistance and secondary leakage reactance. Hardly any current flows through the shunt path due to the very small voltage across it now in the shorted condition. In this circuit, the resistances dominate. hence the circuit current is at high pf.

Dear Anurag, Any open circuit test on any electric machine is designed to test one of the two major materials, which is iron,  used in the machine make-up. ( The no-load test on rotating machines is an equivalent of this test) This test can be interpreted as being conducted to see how effective the transformation of magnetic energy to motive power or transformation to another voltage level ( in transformers) is. 

Hence only magnetic energy is put in when we open circuit the secondary of the transformer and no demand is made by the secondary coils for supplying the load from the source, as reflected current. Hence only magnetism is at work but due to the cyclic nature of the magnetic current hysteresis  and eddy current losses are manifested as power loss or heat. But such losses are designed to be very low since these losses contribute to all day losses. Thus the current drawn from the source, when the transformer is open circuited ( or the machine is on no load) has to subscribe mostly to creating the magnetic field and very little to the losses due to hysteresis and eddy current flow based loss. Clearly this current is lagging with respect to the applied voltage almost 90 degrees. This means a low power factor operation.

At full load or short circuit operation conducted at low voltages, this magnetism related losses are very low. The current drawn from the source is almost fully for supplying the short circuit ( which in the controlled load- only rated current) and hence the power taken from the input is almost fully spent in the copper/ Al. conductors of the machine and the shorting conductors with a very small magnetic component. Hence the current drawn is very much in phase with the applied voltage. Thus almost upf operation is made possible.

Hope the detailing helps!