After months of soaring optimism in the search for materials that offer no resistance to electricity at high temperatures, there is reason to doubt the more extravagant claims. The most extreme had hinted at signs of so-called superconductivity at room temperature and above.

The highest confirmed temperature for a stable superconducting material, however, remains 288 degrees below zero Fahrenheit, or 95 Kelvin -- only slightly better than the 90 K achieved six months ago.

The sudden pessimism has emerged because of widespread failures to reproduce once-promising findings of superconductivity at warmer temperatures. Earlier this month a team from IBM's Almaden Research Center in San Jose, Calif., told the American Chemical Society's annual meeting in New Orleans that it has found two probable causes of the spurious indications of superconductivity at higher temperatures.

One is that there can be changes in the electrical contact between the putative superconducting material and the wires used to test it. A sudden improvement in the contact can look like a drop in resistance. The other probable cause is the heterogeneity of the materials being tested.

The materials are mixtures of yttrium, barium and copper oxide -- powders that are mixed, pressed together and baked at high temperatures. The ingredients arrange themselves into many different crystals of differing sizes and compositions. The electrical signals from such a hodge-podge material cannot be interpreted as straightforwardly as many researchers have been, the IBM scientists said.

The IBM group said it came to its conclusions after getting startlingly promising results, then checking more carefully for sources of possibly misleading data. The group said it could not automatically discount all claims for high-temperature superconductivity but that its findings raised serious doubts that will have to be checked out.