ASSESSMENT OF DEMAND FUNCTION

In the previous chapter, we discussed the demand function, which analyzes how the impact
the price of a product in relation to the quantity of the product demanded by individuals and society.

In this chapter, we will discuss the methods used to obtain data.

demand for decision-making in the business world.

Of course, the hope is that the value of the information obtained should be greater if
compared to the cost used to obtain that information.

The assessment of demand functions is very useful for decision makers to analyze policies.
the company's sales and analyzing the purchasing power of individuals and society. In practice, information
the demand function of a product is not always easy to obtain, therefore to obtain
Information about demand functions, decision-makers can use various ways or methods.
some of them involve conducting in-depth research in the market as well as through approaches
Statistics. Estimating demand relates to how to obtain the parameter values for the function.
relevant requests under the current conditions. This information is important for decision-making at this time
also in evaluating whether the decision has been optimal in the context of the demand situation
at this moment.

A. Assessment and demand forecasting

Before studying demand function estimation further, let us first differentiate the understanding
between estimation or assessment and forecasting or prediction of demand. Estimation
The demand function is the process of determining the coefficient value of a demand function at
a product (current value). Meanwhile, demand forecasting is the process of determining value–
the demand value in the upcoming time period (future value). Values at the present time (current
The value needed to evaluate the optimality of current pricing determination and promotional policies.
as well as to make decisions and company policies. Values for the future
(future value) is needed for production planning, new product development, investment, and
a situation where decisions that must be made have an impact over a long time period.

Estimation and demand forecasting have different objectives. The main objective of estimation
Demand estimation is to evaluate product pricing determination, namely whether the determination
The product price set by the company is in accordance with the capabilities of individuals and society. Forecast
(forecasting) demand is intended as a source of information in planning production
product and product development in the future. The difference between assessment and
Demand forecasting can be explained in the following image:

Year Request
2017 100
2018 150
2019 200
2020 250
2021 ?
2022 ?
Example question:

The data on prices and demand for a product for the periods of 2020 and 2021 is known.

Year Price (P) Demand (Q)

2020 100 40
2021 80 50

Based on the data, calculate the demand function and its interpretation.

Discussion:
P−P1 Q−Q1
=
P2−P1 Q2−Q1

P−100 Q−40
=
80-100 50-40

P−100 Q−40
=
-20 50−40

10P–1.000 = - 20Q + 800

10P = -20Q + 800 + 1.000

10P = -2Q + 1.800

P = -2Q + 180

Q = -0.5P + 90

Q = 90 - 0.5P

Based on that calculation, it can be determined that the demand function is: Q = 90–0.5P

The interpretation of the demand function is as follows:

1. The estimation of the demand function for a product for the year 2020-2021 is:

Q = 90 - 0.5P

The interpretation of the demand function is that when the price of the product in 2021 experienced
a decrease will lead to an increase in product demand by 0.5 in 2021

2. Based on the demand function, the demand for the year can be forecasted.
2022 means that if in 2021 the prices of products are lowered again, then the quantity demanded for the products.
It can be predicted that in 2022 there will be an increase of 0.5.
B. Demand assessment method
According to Arsyad (2011), there are two methods for estimating the demand for a product,
that is:
a) Direct method
The direct method is a method that involves consumers directly.
to individuals and society as a source of information for assessing the demand for a
products. Several demand estimation methods included in the direct methods
among other things:

1) Interviews and surveys

The method of direct demand assessment is by conducting interviews.
for buyers or potential buyers regarding the increase or decrease in quantity
the products they buy if the prices change.
Target groups can be gathered to discuss the issue.
or the questionnaire is directed to a specific sample of buyers. Although
it seems simple, in practice this approach faces many
the obstacles are:
Random skills
The individuals surveyed must represent the overall market so that
the results are not biased. Therefore, the sample must be large enough and use
random method so that the market information is suitable for holding
change plan

Interviewer bias
In this case, the presence of the interviewer can influence feelings.
respondents may provide inaccurate answers.
Interviewer bias often occurs in personal interviews, even
electronic questionnaires such as (email, google form, etc.)

The gap between intention and action

This issue is often referred to as the answer accuracy problem. Consumers
in this case, they do intend to buy the product when interviewed,
but when marketed, something may have changed the intentions and thoughts
the consumer. Finally, the respondent's answer can no longer be obtained.
It is believed that questions asked are confusing or misinterpreted.
or inviting things outside of the consumer's imagination

2) Pass simulation
Another method used to determine consumer response to
Price changes or promotional activities are done by creating a simulated market.
as well as observing the behavior of selected participants in the simulated (artificial) market. How
this is called 'consumer clinic' where it is done by providing a number of
money to the participants and ask them to spend the money on
the artificial environment.
 For different participant groups, prices and demonstrations are set.
different promotions as well. If participants are carefully selected so that they can
representing the market for the product, we can observe after their reaction to
price changes as well as various promotional activities and can draw conclusions
that the entire market will respond to that price change in the same way
The results of this must be observed carefully. There is a possibility of methods used by
participants in spending other people's money will not do so in the same way
in spending their own money. Another possibility is that participants will
choosing a certain product if the price is lowered to make it look like they are
thrifty and responsible consumers.
This method appears to be a data retrieval method with costs.
relative high. This is because we have to provide the products that will be chosen.
The participants and the process take a considerable amount of time.
However, this method can provide useful knowledge.
for us to understand how consumer price awareness and their reactions
generally on changes in certain promotional variables

3) Direct market experiment

This method is similar to the simulation market method, but in
this method the company conducts research on individual behavior in real life
market. This live market experiment involves people in actuality.
being in the actual market situation where they spend their money on
goods and services they need. The company chooses one or more cities, markets
regional, or country and conduct experiments on these 'test markets' design
to seek information on consumer 'acceptance' of products and to identify
the effect of changes in one or more controllable variables on the amount
request

b) Indirect methods
The indirect method is a method that is conducted based on data that has been
collected, then actions are taken to find relationships or correlations
statistically between the independent variable and the dependent variable.
This method uses data obtained from the company, which is then processed accordingly.
statistics to obtain the demand function of a product. Methods to estimate
The demand for a product that is included in the indirect method is regression analysis.
Regression analysis is a statistical technique used to determine how much
the significant influence of an independent variable on a dependent variable
(dependent variable). The following are characteristics of regression analysis, among others:
1. To analyze simple regression, at least two data points are required, namely variable data.
free and data as dependent variables. Variables that affect other variables
referred to as the independent variable, while the variable that is influenced by other variables
referred to as the dependent variable.
 Example question:
The data on the price and demand for a product for the year 2020 is known.
2021
Year Price (P) Demand (Q)
2020 100 40
2021 80 50

Calculate: the interpretation of the demand function.

Discussion:
Based on the data in the table, it can be seen that the price variable (P)
is an independent variable, while the demand variable (Q) is a dependent variable.
bound

2. The data used in the regression analysis is secondary data, such as data
the amount of purchases or sales from year to year or the selling price data of products from
year by year

3. The data used in regression analysis can also be cross-sectional data.

(cross section), time series data and panel data (pooled data). Section data
cross section is data that consists of one object but requires sub
sub-objects related to or that are within the main object
Once upon a time. Time series data is data that consists of a single object
but consists of several time periods, such as daily, monthly, quarterly, and
annual. Panel data (pooled data) is a combination of time series data.
with cross-sectional data. It is called combined data because this data consists of
several objects/sub-objects in several time periods

Example question:
It is known:
Sales Data of Company A Products
Year Harga (P) Sales (Q)
2011 9,000 125
2012 8,000 250
2013 7,000 375
2014 6,000 500
2015 5,000 625
2016 4,000 750
2017 3,000 875
2018 2,000 1,000
Sales Data of Company B Products
Year Price (P) Sales (Q)
2011 9,500 232.5
2012 8,500 347.5
2013 7,500 462.5
2014 6,500 577.5
2015 5,500 692.5
2016 4,500 807.5
2017 3,500 922.5
2018 2,500 1,037.5

Based on the data, determine the time series data and panel data.

Discussion:
Data time series can be shown in the sales data table from company A and
company B, while the panel data is a combination of sales data
Company A and the sales data of Company B for the period of 2011 to 2018.

An example of panel data can be illustrated as follows:

Data panel (pooled data) product sales in the market

Entity Year Price (P) Sales (Q)
Company A 2011 9,000 125
2012 8,000 250
2013 7,000 375
2014 6,000 500
2015 5,000 625
2016 4,000 750
2017 3,000 875
2018 2,000 1,000
Company B 2011 9,500 232.5
2012 8,500 347.5
2013 7,500 462.5
2014 6,500 577.5
2015 5,500 692.5
2016 4,500 807.5
2017 3,500 922.5
2018 2,500 1,037.5
Regression Analysis Application
Mathematically, regression analysis (regression equation) can be expressed as
following:
Y = a + bx
Where:
Y = dependent variable
X = independent variable
A = constant
B = the coefficient of the curve line (the price sensitivity level of the product to the quantity
requested product

The formula to calculate the values of a and b is as follows

∑ −∑ ∑
b=
∑ 2−( ∑ )2
a= ̅ –b ̅
̅ =∑
̅ =∑

The following is an example of manual regression analysis, regression analysis

computerized using SPSS software version 24.0 and spreadsheet
using Microsoft Excel
 Manual regression analysis application
Based on that data, it can be said:
Time Series Data (historical)
Company A

∑ −∑ ∑
b= ∑ 2−( ∑ )2
( 8 x )
19,500,000-(44.000 x 4.500)
b= ( 8 x 284,000,000- (44,000)
) 2

156,000,000 - 198,000,000
b= [Link]−[Link]
-42,000,000
b= 336,000,000
b-0.125

a= ̅ –b ̅
̅ =∑
̅ = 4,500562.50
8
∑
̅ =
̅ = 44,0005.500
8
a = 562.50 - (-0.125 x 5,500)
a = 562,50–(-687,50)
a = 1.250

Based on that calculation, the demand function can be determined.

is Y = 1.250 - -0.125X or Q = 1.250 - 0.125P
Company B

∑ −∑ ∑
b= ∑ 2−( ∑ )2
( 8 x )
25,650,000-(48.000 x 5.080)
b= ( 8 x 330,000,000- (48,000)
) 2

205,200,000−243,840,000
b= [Link]−[Link]
-38,640,000
b= 336,000,000
b -0.115

a= ̅ -b ̅
̅ =∑
̅ = 5.080= 635
8
∑
̅ =
̅ = 48.000 = 6.000
8
a = 635 - (-0.115 x 6000)
a= 605–(- 690 )
a = 1.325
based on that calculation, it can be determined that the demand function Y =
1.325–0.115X or Q = 1.325–0.115P
Data Panel (pooled data)

Combined data from Company A and Company B

∑−∑ ∑
b= ∑ 2−( ∑ )2
)
( 16 x 45,150,000-(92.000 x 9.580)
b= ( 16 x 614,000,000- (92.000)
) 2

722.400.000 −881.360.000
b= [Link] −[Link]
-158,960,000
b= 1,360,000,000
b- 0.116882353
b - 0.117 (rounding)

a= ̅ –b ̅
̅ =∑
̅ = 9.580 598.75
16
∑
̅ =
̅ = 92,0005.750
16
a= 598,75–(-0,116882353 x 5.750)
a= 598,75–(-672,0735)
 a= 1,270.824
a = 1.271 (rounding)
based on that calculation, it can be determined the function
The demand is Y = 1.271 - 0.117X or Q = 1.271 - 0.117P

Computerized regression analysis application

Output data panel with SPSS version 24.0

Model summary
Adjusted R Std. Error of
Model R R Square
Square the Estimate
1 .970a .942 .937 71.725
a. Predictors : (Constant), Harga (P)

ANOVAa
Sum of
Model df Mean
Square F Mr.
Squares
1 Regression 1160992.424 1 1160992.424 225.675 .000
b

Residual 72023.576 14 5144.541

Total 1233016.000 15
a. Dependent Variable : Penjualan (Q)

b. Predictors : (Constant), Harga (P)

Coefficientsa
Unstandardized Standardized
Model Coefficients Coefficients T Mr.
B Standard Error Beta
1 (Constan) 1271.006 48.193 1160992.424 26.373 .000
Harga (P) -.117 .008 5144.541 -15.022 .000
a. Dependent Variable : Penjualan (Q)

The regression analysis calculation using SPSS ver.24.0 can be known

fungsi permintaan pada data gabungan perusahaan A dan perusahaan B (data panel)
is: sales (Q) = 1271.006 - 0.117 Price (P)
The sales demand function (Q) = 1271.006 - 0.117 Price (P) can be illustrated.
as follows:

WORKSHEET 1

Scatterplot of Sales (Q) vs Price (P)

Regression analysis application with Microsoft Excel spreadsheet

C. Coefficient of determination

After understanding the process in regression analysis to estimate the demand for a product, it is
What needs to be understood in regression analysis is the coefficient of determination. The coefficient of determination (R
Square or R squared) or symbolized by 'R' means a contribution
the influence given by independent variables, or in other words, the value
The coefficient of determination or R Square is useful for predicting and seeing how significant
the contribution of the influence given by independent variables simultaneously or together
towards the dependent variable. The formula to calculate the coefficient of determination is as follows:

2
n ∑ XY− ∑ X ∑ Y
R2= [ ]
2
√ [n ∑ X 2-( ∑
X)
]− [n
2
Y -(∑ Y)] ∑ 2

Example question:

Sales Data of Company A

Sales Data of Company B

Sales Data of Company A and B

Calculate the coefficient of determination

Discussion:

Company A
2
n ∑ XY− ∑ X ∑ Y
R 2= [ ]
2 2
√ [n ∑ X 2-( ∑
X)
] [n
2
Y -(∑ Y)] ∑

2
( 8 x 19,500,000-(44.000
) x 4.500)
R 2= [ 20 ( 2
]
√[ (8 x 284.000.000 -44.000]
) ( [ 8 x )3.187.500 - 4.500 ) ( )

2
156,000,000−198,000,000
R 2= [ ]
√ [[Link]−1.936.000 [25,500,000
] - 20,250,000

- 42,000,000 2
R 2= [ ]
√ 336.000.000 5.250.000

-42,000,000 2
R 2= [ ]
√ [Link].000.000

2
R2= [-42,000,000 ]
42,000,000

R 2= 1

Based on the calculation results, the value of the coefficient of determination or R Square is known.
is equal to 1. The value of the coefficient of determination (R Square) is 1 or the same
with 100%. The figure means that the price variable (X) has an influence
towards the sales variable (Y) of 100% at company A

2. Company B
2
n ∑ XY− ∑ X ∑ Y
R 2= [ ]
2 2
√ [n ∑ X 2-( ∑
X)
] [n
2
Y -(∑ Y)] ∑

2
( 8 x 25,650,000-(48.000
) x 5.080)
R 2= [ 20 ( 2
]
√[ (8 x 330.000.000 -48,000]
) ( [ 8 x )3.781.250 - 5.080 ) ( )

2
205,200,000 - 243,840,000
R 2= [ ]
√ [[Link] −[Link] [30,250,000
] - 25,806,400

- 38,640,000 2
R 2= [ ]
√ 336,000,000 x 4,443,600

-38,640,000 2
R 2= [ ]
√ [Link].000.000

2
R2= [−3388.6.64400.0.00000 ]

R 2= 1
 Based on the calculation results, the value of the coefficient of determination or R Square is known.
is equal to 1. The value of the coefficient of determination (R Square) is 1 or the same.
with 100%. The figure indicates that the price variable (X) has an effect.
against the sales variable (Y) of 100% at company B

3. Combined Data of Company A and Company B

2
n ∑ XY− ∑ X ∑ Y
R 2= [ ]
2 2
√ [n ∑ X 2-( ∑
X)
] [n Y −(
2
∑ Y)] ∑

2
( 16 x 45.150.000 -(92.000
) x 9.580)
R 2= [ ]
√[ (16 x 614.000.000 - 92.000]
) ( [ 16)x206,968,750
( - 9.580 ) ( ) 2

2
722,400,000 - 881,360,0000
R 2= [ ]
√ [[Link] −[Link] [111,500,000
] - 91,776,400

-158,960,000 2
R 2= [ ]
√ [Link].000.000

-158,960,000 2
R 2= [ ]
√ [Link].000.000

2
R2= [−116538.7.98600.6.03040 ]

R2 = 0.942

Based on the calculation results, the value of the coefficient of determination or R Square is known.
is 0.942 or equal to 94.20%. This figure indicates that
The price variable (X) has an influence on the sales variable (Y) of 94.20%.
company A and B, while the remainder (100% - 94.20% = 5.80%) is influenced by other variables.
the regression equation is outside this or variables that are not studied

The assessment of the coefficient of determination can also be done through processes other than manual methods.

computerization using statistical software such as SPSS or with

using Microsoft Excel Spreadsheet by using the RSQ function, where
the logic of the RSQ function is:

= RSQ (knows_y’s, known_x’s)

The following is the result of the calculation of the coefficient of determination using
Microsoft Excel Spreadsheet