##
## Call:
## lm(formula = log(wtp) ~ as.factor(marital_status) + time_married +
## level_worry + as.factor(lose_location) + as.factor(lost_previously) +
## log(curr_ring_price) + wow + as.factor(gender) + age_category,
## data = wtp_data_filt)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.11609 -0.51407 -0.00251 0.43673 2.06133
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.12480 3.38076 0.924 0.3645
## as.factor(marital_status)2 4.00783 2.28092 1.757 0.0917 .
## as.factor(marital_status)3 2.66368 2.57104 1.036 0.3105
## time_married 0.47782 0.52193 0.915 0.3690
## level_worry 0.83701 0.67614 1.238 0.2277
## as.factor(lose_location)2 0.73316 0.78591 0.933 0.3602
## as.factor(lose_location)3 0.70691 0.71596 0.987 0.3333
## as.factor(lose_location)4 0.61781 0.92478 0.668 0.5105
## as.factor(lose_location)5 1.34444 0.86217 1.559 0.1320
## as.factor(lost_previously)2 -0.90174 0.52653 -1.713 0.0997 .
## as.factor(lost_previously)3 0.88032 1.55981 0.564 0.5777
## log(curr_ring_price) -0.72740 0.36605 -1.987 0.0584 .
## wow 0.18304 0.08901 2.056 0.0508 .
## as.factor(gender)2 0.69541 0.52405 1.327 0.1970
## age_category -0.49912 0.41971 -1.189 0.2460
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.096 on 24 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.4808, Adjusted R-squared: 0.178
## F-statistic: 1.588 on 14 and 24 DF, p-value: 0.1547
Nothing crazy significant except how high people rate the idea leads them to pay more. There’s a couple other demographics that supposedly are slightly significant but I don’t too much stock in them.
##
## Call:
## lm(formula = wow ~ as.factor(marital_status) + time_married +
## level_worry + as.factor(lose_location) + as.factor(lost_previously) +
## log(curr_ring_price) + wow + as.factor(gender) + age_category,
## data = wtp_data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.9259 -1.4732 0.3151 1.5363 4.5404
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.8753 6.5399 0.745 0.4618
## as.factor(marital_status)2 -1.4965 4.6795 -0.320 0.7513
## as.factor(marital_status)3 1.7365 5.2724 0.329 0.7442
## time_married -1.6941 1.0123 -1.673 0.1046
## level_worry -1.4728 1.4226 -1.035 0.3088
## as.factor(lose_location)2 1.1827 1.6974 0.697 0.4913
## as.factor(lose_location)3 -0.8236 1.5355 -0.536 0.5957
## as.factor(lose_location)4 -0.4454 1.6430 -0.271 0.7882
## as.factor(lose_location)5 -2.3358 1.8124 -1.289 0.2073
## as.factor(lost_previously)2 0.9621 1.0011 0.961 0.3442
## as.factor(lost_previously)3 8.0163 3.0906 2.594 0.0145 *
## log(curr_ring_price) 0.6187 0.6781 0.912 0.3688
## as.factor(gender)2 0.8152 1.1064 0.737 0.4669
## age_category 0.8392 0.8403 0.999 0.3260
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 2.426 on 30 degrees of freedom
## (35 observations deleted due to missingness)
## Multiple R-squared: 0.4348, Adjusted R-squared: 0.1899
## F-statistic: 1.775 on 13 and 30 DF, p-value: 0.09534
Nothings Too significant. I wouldn’t make any claims saying that different demographics rate the idea better or worse. At most I would talk about people who answered “3” to the lost previously question will rate the idea higher on average.
I am assuming you have a fixed cost (overhead, engineering, administrative cost) of roughly $1 Million Dollars, a per unit cost of $200 and that your survey represents 1 Million Adult Americans. Feel free to ask me to change those assumptions.
Really good fit of the data, we got 99% of the data with the model