From the USAF c. 1955 training film No Sweat, there is a technique called 'steer'. By pressing the PTT button for a few seconds, the tower controller is able to determine the pilot's bearing.
I've searched for USAF manuals and Flying Magazine issues from that era, but came up empty.
Looks like a wiki answer would be possible, several of us can provide a part of the full answer. I initiate this one... (before saving your edit, copy your text somewhere, e.g. in your clipboard, as in case of simultaneous updates, you may lose it.)
Feel free to edit and complement!
See also:
Principle: Radiogoniometry, aka direction finding
Available techniques
Determining the bearing of an aircraft from the ground by radio can be done using multiple principles:
With no cooperation from the target: Primary surveillance radar, which means sending a radio wave to the aircraft, and looking for the bearing of the echo returned by the aircraft skin materials.
With automatic cooperation from the target: Secondary surveillance radar, the ground SSR sends an interrogation to the aircraft SSR transponder, which in turn replies by sending pulses which bearing is determined by the ground SSR interrogator.
With manual cooperation from the pilot: The pilot sends a continuous wave using any of the transmitters on-board, e.g. the voice communication transmitter, and the ground station determines the bearing of this signal.
In the documentary video the last technique is used, so let's focus on this one.
Bearing determination techniques
There are several principles which can be used to determine the bearing of the waves from the aircraft:
Peak / null of signal amplitude with a directional antenna privileging reception in one direction. We just turn the antenna until the maximum or minimum signal is found. The maximum gives directly the bearing, the minimum gives the direction +/- 90°. This last option is generally used, as the amplitude rate of change near the null is generally much larger than near the maximum, which increases precision. See How does the ADF null positions find where the NDB is?
Doppler shift: When a radio source moves, the frequency of the source is shifted according to the relative speed between the source and the observer. With some (complex) signal processing, the bearing and trajectory of the source can be extracted.
In 1955, only the first technique was used.
By using two distant stations, two intersecting bearings are available, the distance to the source can also be determined. The technique is known as triangulation (the two stations and the source being the three apexes of the triangle).
Receiver on the ground: Antenna
We know the principle to use: The pilot triggers a continuous wave transmission the time for the ground station to determine the maximum or null direction.
The first generation of direction finders used rotating antennas. Using an antenna with an asymmetrical sensitivity, the voltage at the antenna output is dependent on the direction of the source, according to the antenna radiation pattern (which is also the reception pattern). The antenna can be rotated until the maximum or minimum is found.
A first improvement has been to use the principle of interferometry. With two or more antennas. The same signal reaches the antennas with a difference of phase due to the different distances traveled, unless the source is perpendicular to the antenna array plane. With this technique, the antennas mustn't be directional, they can be more simple: a vertical pole. A system using four antennas, creating two reception directions (North-South and West-East), has been deployed in UK on 1920-1930 (the Adcock antenna network and the Huff-Duff/HD-DF receiver)).
Adcock, Huff-Duff/HD-DF and Bellini-Tosi/BT-DF
The principle of the Huff-Duff was to use the N-S and W-E antenna outputs to drive the electronic beam of an oscilloscope. An oscilloscope is (was) a cathodic ray tube, with a source of electrons and two pairs of deflecting electrodes, one to determine the horizontal position of the electron beam on the front face, and one to determine its vertical position. By ensuring the displacements were a sine/cosine transformation of the Adcock array output, the beam was directed to a spot indicating the bearing of the source.
With this system, it is not required to turn the antennas, and a bearing can be determined instantaneously.
A similar direction finder was also built in the US by Marconi: Bellini–Tosi direction finder (B–T or BTDF) using a loop antenna and two vertical antennas (sense antennas) to remove the front-back ambiguity. The output of the antennas are quite the same than the Adcock system, except being not referenced to the North direction (so technically speaking, the Adcock system provides an azimuth, and the B-T system provides a bearing).
Ground direction finder for US Navy
...
From ground direction finder to airborne direction finder
...
Ground radiogoniometry circa 1955 in the US
...
Time to get a bearing
...
Ground radiogoniometry today
...