Issues of validity in openly ideological research: Between a rock and a soft place