Research in Astronomy and Astrophysics LETTERS A critique of supernova data analysis in cosmology

Abstract Observational astronomy has shown significant growth over the last decade and has made important contributions to cosmology. A major paradigm shift in cos-mology was brought about by observations of Type Ia supernovae. The notion that the universe is accelerating has led to several theoretical challenges. Unfortunately, although high-quality supernovae data-sets are being produced, their statistical anal-ysis leaves much to be desired. Instead of using the data to directly test the model, several studies seem to concentrate on assuming the model to be correct and limiting themselves to estimating model parameters and internal errors. As shown here, the important purpose of testing a cosmological theory is thereby vitiated. Key words: cosmology: observations — supernovae Ia: general Observations play a crucial role in all branches of science. Moreover, they are more important in cosmology where, on one hand, the events are non-repeatable and, on the other hand, the theo-retical side is more speculative than laboratory physics, requiring guidelines from the observations by confronting them. The power of observations in cosmology is clear from the observations of supernovae (SNe) of type Ia which dramatically changed, about a decade ago, the then standar