The main goals of the present study were to verify whether the decision-making processes of choosing attractive females and choosing money have differential neurophysiological mechanisms and, if so, to identify different processing stages in the decision-making between attractive female rewards and money rewards. In order to realize these goals, we devised a novel task of decision between viewing an attractive female clearly and gaining a certain amount of money for healthy heterosexual male volunteers and recorded their brain EEG responses. Our experimental paradigm and procedure were elaborately designed so that male subjects would truly express their evaluation and decisions. Using this framework, we obtained several discoveries. First, the response time (RT) of choosing to view attractive females was significantly longer than choosing to gain money. The reason might be that choosing to gain money represented an optimal action (if a participant chose money, 1 fen
could afford him to use Internet to browse attractive females for 1.5 minute; if he chose an attractive female, he could view the attractive female for only 1.5 second), and so a subject generally hesitated more before he chose to view an attractive female in clear than to gain money. Second, the type of attractive female pictures influenced the response time of decision-making in males. Specifically, behavioral results indicated that the RT for task S was longer than task B. The reason might be that the information of sexy females captured a male's attention easier and hindered their performance (delayed reaction time) in the decision-making task. These results were consistent with the notion of evolutionary theory that the erotic signal can easily attract males' attention, and this played an important role in mating and reproducing decisions for human beings 
. Another possible reason was, the CEs for sexy female images were generally larger than those for beautiful ones, and so both options in task S (sexy images and higher money amount) were more attractive than those in task B (beautiful images and lower money amount), and so subjects would feel more conflicted in task S, which lengthened the reaction time in task S.
Scalp ERP analysis yielded several new interesting findings: the positivity in the time interval of 290–340 ms (P2) in task B was significantly larger than task S. Moreover, in the time window of 340–390 ms, the decision-making in task S elicited a greater negative component (N2) than in task B. Finally, the decision-making in task S elicited a greater late positive component (LPC) than in task B during the 550–1000 ms time window, and choosing an attractive female evoked a greater LPC than choosing money. We discuss the implications of these findings in terms of the association between ERP components and decision-making as follows.
The decision-making in task B elicited a more positive ERP deflection (P2) than in task S following onset of an attractive female stimulus during 290–340 ms. The P2 has been often studied in visual search 
, language processes 
, and memory recognition tasks 
. Multiple evidence demonstrated that the P2 as a part of cognitive matching system, might reflect general neural processes that compared an external visual input with an internal memorial information 
. In the present study, beautiful females were characterized with their attractive faces, while sexy females were featured by their erotic postures, and participants might focus on females' faces and postures during task B and task S respectively. Due to the complexity and specialization of human brains processing human faces, the beautiful females may require more cognitive resources in perceptual-matching process than sexy females, thus elicited larger P2. This result could not be explained by sexy arousal, because if this was the case, the sexy females would have elicited larger P2, opposite to the actual result. This result indicated that during the early processing of the decision-making, the perception process may be different for beautiful and sexy female tasks.
The decision-making in task S demonstrated a greater negativity (N2) at 340–390 ms than in task B. Previous work indicated that the N2, which was regarded as a component of mismatch negativity or mismatch detector 
, often appeared in go/no-go tasks 
and reflected a kind of cognitive inhibitory mechanism of controlling incorrect response 
. However, a recent study provided evidence that fought against the opinion of N2 representing response-inhibition, and their results showed that N2 should represent conflict monitoring 
and previous explanations of the N2 as indexing response inhibition might need to be revised. Moreover, other studies also suggested that the N2 might reflect conflict monitoring on correct response trials 
and the high degree of response conflict might derive from the competition between the execution and the inhibition of a single response in go/no-go tasks 
. In the current study, when participants confronted the choices of a beautiful female and money or a sexy female and money, they might have different degrees of response conflicts in the two tasks. The response time in task S was longer than in task B and the money amount of the cash equivlent in task S was generally higher than in task B. Thus, the cognitive response conflict (wanting both money and an attractive female) in task S might be larger than that in task B. Task S elicited a greater negativity (N2) than task B, might be related to larger conflict monitoring.
During the 550–1000 ms time window, choices in task S were accompanied by higher positive amplitudes (LPC) than those in task B. LPC was generally considered to relate to valuation processes 
. Different valuation processes may involve different subjective values. Mainly, there are two kinds of subjective values: decision value (the value that people assign to the options, and on which they base their decisions) and experienced value (the value that people experience when they get and consume something) 
. According to the above definition and our experimental procedure, we regarded the value in the current consideration as decision value rather than experienced value. In our experiment, cash equivalents for sexy female images were generally higher than those for beautiful ones, indicating that sexy female images had higher decision value than beautiful ones, and so sexy images induced higher LPC than beautiful ones, as observed in the current experiment.
Another different but related explanation is as follows. Previous studies have found that P600 and other late positive ERP components were involved in erotic content in both males and females 
. A recent study used unique design directly to compare the effects of erotic and non-erotic stimuli 
, and their results showed that erotic pictures yielded significantly higher positive slow waves (PSW) between 500 and 750 ms than sport stimuli with comparable arousal and valence levels. It was suggested that the larger arousal value and positive valence of erotic stimuli, both of which were evolutionarily relevant, were the core element of ‘motivated attention’ and played an important role during motivation-directed behaviour 
. Thus, we inferred that evolutionary relevance might be a factor that accounted for our observed results. Specifically, sexy stimuli can be regarded as more relevant to survival and reproduction of human beings than beautiful stimuli 
, and so participants put higher subjective values to sexy females than beautiful females, and thus the former induced larger LPC than the latter.
In addition, during the 550–1000 ms time window, the positivity of choosing an attractive female picture was significantly larger than choosing a certain amount of money (cash equivalent). Imagine a male faced a vague image of an attractive female and needed to decide to view its clear version or to gain a fixed amount of money which was roughly as attractive as this kind of images. If he assigned relatively higher subjective decision value to this vague image, then (i) the corresponding ERP amplitude would be relatively larger and (ii) he would choose to view its clear version, because the decision value of this image was slightly higher than that of the fixed amount of money. If he assigned relatively lower subjective decision value to this vague image, then (i) the corresponding ERP amplitude would be relatively smaller and (ii) he would choose to gain a fixed amount of money, because the decision value of this image was slightly lower than that of the fixed amount of money. Therefore, choosing images of attractive females were preceded by larger amplitude of LPC, as observed in the current experiment.
Another possible explanation is, relative to the money amount, female pictures, no matter sexy ones or beautiful ones, were more directly related to humans' survival and reproduction, and their relevant cognitive and affective responses occurred earlier in the human beings' evolution. First, according to previous studies 
, erotic stimuli were especially relevant to human's survival and reproduction. Second, a person with a more beautiful face was often supposed to be friendlier 
, and so a more beautiful face could be socially more rewarding. In addition, when making social decisions, humans might take the degree of beauty of faces into account. Both sexy and beautiful female images could begin to play an important role in human life in the early stage of evolution. Finally, compared with attractive females, money occurred quite late in the human evolutionary history, and the rewarding value of money was indirect (money itself cannot satisfy your appetite or any other kind of desire). In a word, as a kind of rewards, attractive female images were more direct and evolutionarily deeper than money, and so could induce stronger neural responses than their cash equivalents. When the subjects chose attractive female pictures over cash equivalent, they anticipated more direct rewards and so had stronger LPC.
In conclusion, the current study delineated the neural processes of the decision-making between attractive females and money in male subjects using ERP technique. Our findings clearly indicated that in the time window of 290–340 ms, task B evoked a larger P2 than task S, which may be related to a process of matching external perception with an internal representation or expectation in the early stage of the processing; in the time window of 340–390 ms, task S evoked a larger N2 than task B, which may relate to the response conflict and cognitive monitoring; finally, during the 550–1000 ms time window, decision-makings in task S were accompanied by significantly higher positive ERP components than those in task B, reflecting sexy female images may have higher decision value than beautiful ones, which may be because the former is evolutionarily more significant; more importantly, choosing an attractive female was accompanied by a larger late positive component than choosing money and this is possibly because the decision values of the chosen attractive female images were generally higher than those of the un-chosen ones, or because attractive female images are more direct and evolutionarily earlier rewards than money amounts, and this can influence the subjective evaluation and expectation process. Our study represents the first attempt to understand the neural processes of decisions between attractive females (a typical example of primary rewards) and money (a typical example of secondary rewards).