Hi, I have a question from my homework. Parts to this question are in Bold.

Aventurier is interested in expanding its internet footprint in part with more direct consumer sales though its "Build-Your-Own-Board" (BYOB) page that will be linked to Aventurier's homepage. The company is considering enhancing its BYOB page with (1) "swishing sounds" of snowboards racing down the slopes, played either softly or moderately loudly in the background, and/or (2) videos of artisans crafting details on boards, small videos that would be shown one at a time on the right side of the screen.

MarketWise therefore conducted a 2 (Swishing-Sound-Volume: soft vs. moderate) x 2 (Artisanal-Video: absent vs. present) between-subjects factorial experiment in real time on Aventurier's beta test of its BYOB area.

The 4,000 consumers clicking on the BYOB link during the two-week test period were randomly assigned to one of four experimental conditions (roughly 1,000 assigned to each condition; N = 4,000). The key outcome measure (DV) was sale (no, yes), which was summarized as each condition's sales percentages. Marketwise also drilled down into consumer thoughts and opinions by giving consumers the option to take a short survey after they visited the BYOB-area, which roughly 50% of the visitors did.

Marketwise is scheduled to send Aventurier its final report in two weeks, but a friend at Marketwise just sent your boss the per-condition sales percentages ahead of time because they know your boss is impatient for the results. Unfortunately, your boss is dismayed because the results do not seem to surface much. Here are the sales percentages and number of consumers per condition: Soft-Sounds-No-Video = 4.5% (n = 960), Soft-Sounds-Video = 4.0% (n = 1,050), Moderate-Sounds-No-Video = 4.0% (n = 950), Moderate-Sounds-Video = 3.5% (n = 940).

For the following Parts a-d, assume that the sales differences smaller than 1.0% are non-significant at p = .18, but that those greater than or equal to 1.0% are significant at p < .05.

CRITICAL: To simplify, when calculating means across conditions for the following questions, please do NOT weigh the conditions by their cell sizes, but simply add their means and divide by the number of conditions. For example, mean sales in the Soft-Swishing-Sounds condition for the following analyses are simplified to be (4.5% + 4.0%)/2 = 4.25%, where 4.5% is the mean in the Soft-Sounds-No-Video condition, and 4.0% is the mean in the Soft-Sounds-Video condition. Cell-size weightings would change little, and given that our goal here is simply to make sure we all understand main and interaction effects conceptually, we simplify the cross-condition-mean calculations for the following questions.

a. Swishing-Sound-Volume Main Effect: What are the two means and the difference tested to assess this main effect, and is this main effect statistically significant?

b. Artisanal-Video Main Effect: What are the two means and the difference tested to assess this main effect, and is this main effect statistically significant?

c. Swishing-Sound-Volume by Artisanal-Video Interaction: What are two of the mean-difference pairs and their difference that is tested to assess this interaction, and is this interaction statistically significant?

d. If your boss is considering only the experiment's four options, which would you recommend and why?