When I started observing variable stars in January 1999, I obtained my first results by means of a small 8x30 binoculars (Carl Zeiss Jena). The first objects have been U Mon and delta Cep. I started using the Argelander approach of estimating star brightnesses.

For a star with a long period (e.g. U Mon), you can obtain nice light curves just by making observations in every night. The estimates are then plotted vs. time (given as the julian date). 

Cepheids such as delta Cep or zeta Gem have a fairly short period of only a few days. In order to obtain a continuous light curve, the individual estimates have to be reduced, that is, taking the known period to calculating back all individual estimates to a narrow period of time. This is called "folding of the light curve".

Most other short period variables require for another observation approach. RR lyrae stars and most eclipsing binaries are observed by performing 5 to 12 estimations per hour, depending on how short the period really is. The Argelander approach is useful especially here, because no absolute star brightnesses are needed.