An extensive literature has studied ambiguity aversion in economic decision making, and how ambiguity aversion can account for empirically observed violations of expected utility-based theories. Almost all relevant applied models presume a general dislike of ambiguity. In this paper, we provide a systematic experimental assessment of ambiguity attitudes in different likelihood ranges and in the gain domain, the loss Domain and with mixed outcomes. We draw on a unified framework with more than 500 participants and find that ambiguity aversion is the exception, not the rule. We replicate the usual finding of ambiguity aversion for moderate likelihood gains. However, when introducing losses or lower likelihoods, we observe either ambiguity neutrality or even ambiguity seeking behavior. Our results are robust to different elicitation procedures.