It has been shown from an evaluation of the inverse reading of the dosemeter (1/M) against the inverse of the polarizing voltage (1/V), obtained with a number of commercially available ionization chambers, using dose per pulse values between 0.16 and 5 mGy, that a linear relationship between the recombination correction factor kS and dose per pulse (DPP) can be found. At dose per pulse values above 1 mGy the method of a general equation with coefficients dependent on the chamber type gives more accurate results than the Boag method. This method was already proposed by Burns and McEwen (1998, Phys. Med. Biol. 43 2033) and avoids comprehensive and time-consuming measurements of Jaffé plots which are a prerequisite for the application of the multi-voltage analysis (MVA) or the two-voltage analysis (TVA). We evaluated and verified the response of ionization chambers on the recombination effect in pulsed accelerator beams for both photons and electrons. Our main conclusions are: (1) The correction factor k(S) depends only on the DPP and the chamber type. There is no influence of radiation type and energy. (2) For all the chambers investigated there is a linear relationship between kS and DPP up to 5 mGy/pulse, and for two chambers we could show linearity up to 40 mGy/pulse. (3) A general formalism, such as that of Boag, characterizes chambers exclusively by the distance of the electrodes and gives a trend for the correction factor, and therefore (4) a general formalism has to reflect the influence of the chamber construction on the recombination by the introduction of chamber-type dependent coefficients.