Using reliability measures to analyze survey items

Consider the executives at a TV studio. They are concerned about the number of people who watch the shows they produce. Recently, they needed to decide whether to continue producing a popular show. It had run for several years, and some of the actors, directors, and screenwriters wanted more money or wanted to move on to other projects. The studio was willing to spend more money on the show but only if it would continue to be popular, so they conducted a survey of randomly selected television viewers. Realizing that there are many reasons why a person may or may not watch a particular show, seven items were constructed for this survey. The questions were of the form, Would you watch this show next season if:?

Answers from 906 respondents can be found in the tv-survey.sav data file. See the topic Sample Files for more information. Yes answers are coded with a 1; No answers, with a 0. Use Reliability Analysis to determine to what degree the makers of the survey were successful in constructing questions that measure a person's opinion.