I am a bit confused between packet switching and circuit switching usage in telephones. If there is a loss of packet in packet switching or the packets may arrive in different order why is it still used for telephone and video calls etc.? Why not use circuit switching when all the data is transferred without any loss and in order? I hope I made myself clear.
This may be a bit of a cynical answer, but telecommunications companies are not in business to provide high fidelity audio, they're in business to make money.
In other words, it's a matter of economics. Packet switching is more economical and flexible, which makes telecom companies more profitable. The reason these companies have invested so much in packet switching hardware is that they can offer more services for lower cost.
The fact is, the occasional out of order or lost packet does not degrade voice quality enough to make a noticeable difference. Audio codecs can compensate for lost data so that it's undetectable by most people.
So higher quality voice does not justify the higher costs. Very few customers would be willing to pay the extra cost, or even notice the difference.
If there is a loss of packet in packet switching or the packets may arrive in different order why is it still used for telephone and video calls etc.?
Packet switching is much more efficient and thus several times(!) cheaper than circuit switching. It also scales much more easily. After all, as a business you want to earn money (as Ron T. has already pointed out) and the basic approach is to reduce your cost.
Additionally, packet loss is possible but if the rate is very low it's acceptable almost everywhere - especially when the network cost is only a fraction or - put the other way - you get several times the performance for the money.
With a good network design (topology, redundancy, QoS, ...), packet loss is (far) below 1 ppm. Voice and video applications can handle 1000 ppm loss very well.