What does it take to win Australian Idol?

The effects of singing ability and appearance on the likelihood of receiving votes

BACKGROUND TO THE STUDY AND HYPOTHESES

The past decade has seen an explosion of "unscripted" reality shows on television stations around the world. Talent competitions, in particular, have become very popular, covering various skills such as singing (The Voice, Australian Idol), cooking (My Kitchen Rules, Masterchef), modeling (Australia's Next Top Model), and business (The Apprentice). Each show features a panel of expert judges who assess the contestants' performances. However, many shows also rely on viewers to determine which contestants should advance in the competition and which ones should go home.

The Australian public's votes can shape the outcome of these talent competitions in powerful ways. However, little is known about the process by which viewers decide to vote for a particular contestant. In this study we consider what it takes to win a show such as Australian Idol, focusing on two contestant characteristics that determine the likelihood of receiving votes. This question is of great interest to the producers of such shows (who want to verify the validity of the voting process), as well as to the contestants (who want to improve their chances of winning the competition).

One factor that is likely to determine people's votes is their assessment of the skills relevant to the competition. Research indicates that people who are asked to judge others on a particular dimension (e.g., public speaking ability) systematically assess the positive and negative aspects of the individual's performance in order to form this judgment (Horner & Vines, 2014; Ohbuchi, Kameda, & Agarie, 2001). In addition, game theory proposes that evaluations made in a competitive context are based on an assessment of the relevant skill-set (Miller, 1993; Stevens, Ng, & Sandler, 2012). This suggests that viewers will be more likely to vote for a contestant on Australian Idol if they believe that s/he has great talent as a singer.

Another factor that is likely to determine people's votes is their assessment of contestants' appearance. A long line of research has demonstrated that individuals form more positive attitudes towards an attractive person compared to an unattractive person (e.g., Rafferty, 2005). However, another feature that could be important is the extent to which a person's appearance is normal or quirky (i.e., distinctive). People who have a normal appearance may not receive as much attention as those who have a quirky appearance (Credlin & Smith, 2011; Smith, Credlin & Appleby, 2013), and people with a distinctive appearance may be inferred to be more extraverted (Nauman, Collins, Bauer, & Lipp, 2009). Thus, it is possible that viewers will be more likely to vote for a contestant on Australian Idol if s/he has a quirky appearance, rather than a normal one. 2/7

In addition, there may be an interplay between these two factors. Specifically, when a contestant has a quirky appearance, viewers' judgments of the contestant's singing ability may be less important in determining their voting decisions. On the other hand, when a contestant has a normal appearance, viewers may base their voting decisions on judgments of relevant skills such as singing ability. Although these propositions seem plausible, they have not been empirically tested.

In this experiment we examined how a female contestant's singing ability and appearance operate together to affect viewers' likelihood of voting for her. By using standardized videos of a hypothetical contestant, we were able to independently manipulate the two factors of interest (contestant appearance and singing ability), while controlling all other aspects of the performance. We predicted that, overall, individuals would be more likely to vote for a contestant who had a quirky appearance, compared to a contestant who had a normal appearance. We also predicted that, overall, individuals would be more likely to vote for a contestant with better singing ability. Finally, we expected that the hypothesized effect of singing ability would be limited to cases in which the contestant's appearance was normal.

INSTRUCTIONS FOR ASSIGNMENT

The aim of this assignment is to consolidate your understanding of factorial designs and their analysis, and to give you practice in writing up the results in the appropriate format. Using the data provided below, your task is to complete this research report by writing the Method, Results and Discussion sections.

Note that you are not expected to read the references cited on pages 1-2, or any other references on judging ability, quirky vs normal appearance, or voting in reality TV shows. Your report should be written as it would for publication in an APA psychology journal (but again, references are not required). However, you will not need to provide a title page, Abstract, Introduction or References section. Therefore, you need to do the following:

Preliminary Tasks

- _Conduct appropriate statistical analyses to test the effectiveness of the experimental manipulation of singing ability (see Table, Results for the Manipulation Check).

- _You should conduct all the relevant analyses using SPSS, and append the SPSS syntax to your report.

- _You are required to report follow-up tests and effect sizes for the manipulation check. However, for this variable you can report the partial eta squared (η2P) and pairwise comparisons from the SPSS output (so that you do not have to do hand calculations). Note: Ideally, manipulating singing ability would produce an effect of singing ability at all levels of appearance type.

- _Conduct appropriate statistical analyses to test the focal predictions about the effects of singing ability and appearance type on the dependent measure (see Table 2).

- _Use SPSS to test the omnibus effects and any appropriate follow-up tests. For follow-up tests involving the comparison of groups, if the hypotheses permit you can report pairwise comparisons from the SPSS output rather than using hand-calculated linear contrasts. Do not forget to append the SPSS syntax.

- _Report an effect size for every F test. We will imagine that you have a statistically-savvy advisor or journal editor who rejects partial effect size measures, so you must use either η2 or ω2. N.B. you are not required to append these hand calculations.