In an experiment described in the article “The Impact of SMS Advertising on Members of a Virtual Community” (J. Advert. Res. 2008: 363–374), researchers worked with an online gaming forum to send out messages advertising a deal on Subway sandwiches. The message was varied in two ways: the apparent spokesperson (either Subway or “Nik,” a made-up forum member) and the language used (full English or texting shorthand). Forum members were randomly assigned to receive one of the four possible messages (spokesperson-language pairs); recipients were later surveyed about their views on the advertisement (attractiveness, credibility, etc.). Assume the researchers obtained K = 12 survey responses for each message (which approximates the actual study results).

a. Credibility ratings were translated into normalized scores for each participant, with positive values meaning the message was more credible (zero neutral, negative less credible). Mean normalized credibility scores appear below.

Construct an interaction plot, and describe what you see.
b. With A = spokesperson and B = language, sums of squares include SSA = 4.2905, SSB = 0.2926, SSAB = 0.0818, and SSE = 36.5930. Perform a complete two-factor ANOVA, testing each of the three possible effects hypotheses at the .05 level. Explain what you discover. 

c. Researchers also the measured “purchase intention” of each subject, with higher numbers indicating a greater likelihood of buying a Subway sandwich. Use the information below to create an interaction plot and perform a two-factor ANOVA as in parts (a)–(b). Again, explain your findings.

Transcribed Image Text:

Credibility Subway Nik Full English .2729 -.4076 Texting shorthand .3465 -.1689