Weak interaction processes play an essential role in many astrophysical scenarios. In many cases their description involves a knowledge of the properties of many nuclei that have to be theoretically modeled since experimental information is not available. It is essential that sufficient experimental data are available to test, constrain and guide the theoretical models. This paper concerns the determination of the absolute value of the Gamow-Teller transition strength B(GT) in fp-shell nuclei. These nuclei are important e.g. in neutrino induced reactions at the violent core-collapse stage of type II supernovae. Experimental B(GT) values can be obtained from charge-exchange (CE) reactions on stable nuclei or from beta decay since they involve the same operator. These two studies are to a large extent complementary, each having their own advantages and disadvantages. Assuming isospin symmetry one can combine the results from both types of study on mirror nuclei to determine absolute B(GT) values. This method, the so-called 'merged analysis' is very powerful since it allows B(GT) determinations for GT transitions starting from very unstable nuclei if the beta-decay T1/2 and the Q-beta value are precisely known and if the relative B(GT) values are studied in CE reactions for the mirror nucleus. In order to determine the degree of precision of this method a campaign has been launched at several different laboratories to study the decays of Tz=-1 and Tz=-2 nuclei in the fp shell with the aim of comparing the results with the results from CE experiments on the corresponding Tz=+1 and Tz=+2 mirror nuclei carried out at RCNP (Japan). Recent experimental developments have made this comparison effective. On the one hand there has been a substantial improvement in resolution in studies of (3He,t) CE reactions at RCNP. On the other hand the use of fragment separators to create and identify short-lived nuclear species has improved experimental accessibility to the beta decays of exotic nuclei (all nuclei studied here have T1/2 of the order of 100 ms). We will present here a summary of the results with special emphasis on the precise beta decay studies obtained at the GSI and GANIL facilities. The experiments include DSSD implantation- beta- proton detectors and Ge-gamma arrays. The experiments resulted in T1/2 determinations for more than 20 nuclei with one to two orders-of-magnitude improvement in accuracy and absolute branching ratios to excited states up to 5 MeV excitation in nuclei where typically only the branching to the lowest state was known previously. A special effort has been made to determine the strength of the g.s to g.s beta decay branch, which is notoriously difficult to measure.< Réduire