In: Math
A furniture store has maintained monthly sales records for the past 20 months, with the results shown below.
Month |
Sales |
1 |
2360 |
2 |
1820 |
3 |
1760 |
4 |
1560 |
5 |
1950 |
6 |
1950 |
7 |
3360 |
8 |
1740 |
9 |
3780 |
10 |
2400 |
11 |
2160 |
12 |
2760 |
13 |
3570 |
14 |
2820 |
15 |
2800 |
16 |
1890 |
17 |
2500 |
18 |
3630 |
19 |
2530 |
20 |
3270 |
Assume you have determined there is NO SEASONALITY in this time series. Therefore, you want to fit a linear trend model (that is, trend only) to the data.
Calculate the linear trend equation. (Round coefficients to the nearest whole number.)
y Subscript t Baseline equalsyt=nothing+
nothing*t
What are the test statistic and p-value to test for a significant trend. Round both to two decimal places.
T = nothing
p-value = nothing
Is the trend significant using a 10% significance level?
Yes
No
What is the value of R-squared? (Round to two decimals.)
nothing
Forecast the sales for the next month (t = 21). (Round to the nearest whole number.)
Upper F 21 equalsF21=nothing
Based on the R-squared value, how confident are you in this forecast? (That is, how accurate do you think the forecasts will be?)
A.
Not confident at all because the R-squared value is so low
B.
Very confident because the R-squared value is high
C.
Somewhat confident because the R-squared value is moderate (not extremely high but not extemely low)
x | y | (x-x̅)² | (y-ȳ)² | (x-x̅)(y-ȳ) |
1 | 2360 | 90.25 | 29070.25 | 1619.75 |
2 | 1820 | 72.25 | 504810.25 | 6039.25 |
3 | 1760 | 56.25 | 593670.25 | 5778.75 |
4 | 1560 | 42.25 | 941870.25 | 6308.25 |
5 | 1950 | 30.25 | 336980.25 | 3192.75 |
6 | 1950 | 20.25 | 336980.25 | 2612.25 |
7 | 3360 | 12.25 | 688070.25 | -2903.25 |
8 | 1740 | 6.25 | 624890.25 | 1976.25 |
9 | 3780 | 2.25 | 1561250.25 | -1874.25 |
10 | 2400 | 0.25 | 17030.25 | 65.25 |
11 | 2160 | 0.25 | 137270.25 | -185.25 |
12 | 2760 | 2.25 | 52670.25 | 344.25 |
13 | 3570 | 6.25 | 1080560.25 | 2598.75 |
14 | 2820 | 12.25 | 83810.25 | 1013.25 |
15 | 2800 | 20.25 | 72630.25 | 1212.75 |
16 | 1890 | 30.25 | 410240.25 | -3522.75 |
17 | 2500 | 42.25 | 930.25 | -198.25 |
18 | 3630 | 56.25 | 1208900.25 | 8246.25 |
19 | 2530 | 72.25 | 0.25 | -4.25 |
20 | 3270 | 90.25 | 546860.25 | 7025.25 |
ΣX | ΣY | Σ(x-x̅)² | Σ(y-ȳ)² | Σ(x-x̅)(y-ȳ) | |
total sum | 210 | 50610 | 665 | 9228495.0 | 39345.00 |
mean | 10.50 | 2530.50 | SSxx | SSyy | SSxy |
sample size , n = 20
here, x̅ = Σx / n= 10.50 ,
ȳ = Σy/n =
2530.50
SSxx = Σ(x-x̅)² = 665.0000
SSxy= Σ(x-x̅)(y-ȳ) = 39345.0
estimated slope , ß1 = SSxy/SSxx = 39345.0
/ 665.000 = 59.16541
intercept, ß0 = y̅-ß1* x̄ =
1909.26316
so, regression line is Ŷ = 1909
+ 59*x
-----------
estimated std error of slope =Se(ß1) = Se/√Sxx =
619.168 /√ 665.00 =
24.0103
t stat = estimated slope/std error =ß1 /Se(ß1) =
59.1654 / 24.0103
= 2.46
p-value = 0.024
Is the trend significant using a 10% significance level?
Yes
---------------
R² = (Sxy)²/(Sx.Sy) = 0.25
Predicted Y at X= 21 is
F21 = 1909.2632
+ 59.1654 * 21
= 3152
-----------
Not confident at all because the R-squared value is so low