DZNeP-Daughter Has Proved The Most Up-To-Date Strategy . . . Learning To Make A Fortune Completely From Scratch

Although the outcome was not provided directly to a participant, the announcement varied between the conditions. In the positive condition the robot hinted approval of what the participant said during the interview. In the negative condition it was not particularly pleased with a participant's responses suggesting them to consider applying elsewhere. At that point participants were asked to fill the questionnaires for the third time. This time multiple dummy questions regarding the interview were included. Interaction III was identical as Interaction II, but the CVs, job positions and questions asked by the robot were different. Participants were permitted to answer each of the questions freely and we did not measure the duration of interactions. The whole procedure took approximately 1 h. 2.4. Measurements In the experiment we have used several questionnaires and the BIAT (Sriram and Greenwald, 2009) as dependent measures. We explicitly measured the robots' perceived eeriness and anthropomorphism on 5-point Likert scales derived from Ho and MacDorman (2010). Moreover, likeability was measured using the corresponding Godspeed scale from Bartneck et al. (2009b) (range 1�C5). In order to establish the relationship between multi-dimensional anthropomorphism and the uncanny valley we have measured 2 dimensions of anthropomorphism: HN and HU on scales developed by Haslam et al. (2009). Both dimensions had 10 items and were measured on a scale from 1 (not at all) to 7 (very much) (e.g., ��The Robo is�� shallow��). This experiment is part of a bigger study that involved additional self-report scales that were collected at the same time and are not reported here. We used a validated version of likeability scale in Japanese. Perceived eeriness, anthropomorphism, HN and HU were available only in English. Therefore, we conducted a back-translation process to obtain their Japanese versions. We calculated reliability of each scale separately for each interaction round using Cronbach's ��. According to Nunnally (1978) Cronbach's �� > 0.6 is acceptable for newly developed scales for research purposes. Based on this threshold, all the scales, apart from HU were adequately reliable. The lowest Cronbach's �� BI 6727 solubility dmso values during any of the three measurements were as follows: likeability �� = 0.83, perceived eeriness �� = 0.62, anthropomorphism �� = 0.88, HN �� = 0.65 and HU �� = 0.54. Low reliability of HU scale indicates that the results for this scale should be interpreted with great caution. Furthermore, we used BIAT (Sriram and Greenwald, 2009) as a computer-based implicit measurement tool of eeriness. BIATs involve participants classifying series of words into superordinate categories. The task involved combining concept classification (��Robo�� vs. ��Human��) with an attribute classification (��Eeriness�� vs. ��Non-eeriness��). We were interested in measuring the strength of association between ��Robo�� and ��Eeriness.