Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

Prohibition in ACBC

i'm currently working on the study were we need keep Price increments to zero for an level i.e. keep base price for that level  let say level 5 of the attribute 2). how we set this condition in the ACBC conjoint exercise. Please help and suggest.
asked Aug 21, 2019 by Milan Sanathra (470 points)

1 Answer

0 votes
There are two ways to use a Price attribute in ACBC: 1) Summed Pricing, or 2) Standard Price Attribute.  I'm going to assume you are using Summed Pricing because you are talking about price increments per level.

First, let's make sure you are talking about the Summed Pricing approach.  In the Summed Pricing Approach, you add an item to your Attribute List for Price.  Then, on the Attributes tab in the ACBC settings, you click in the "List of Levels" column that the Price attribute is "Summed Pricing Attribute".   Next, on the Pricing tab you specify a Base Price, a percent random shock (usually +30% to -30%) and other settings that govern price (any decimals of precision to show, any rounding, etc.).  Per each non-price attribute, you can specify that each attribute level adds or subtracts a certain price amount from the base price.  

Now then, you don't need to add or subtract any price amount from the base price for an entire attribute or any level of an attribute.  For the attribute level in which no adjustment to the base price is to happen, when you use the drop-down control on the Pricing tab to select that attribute, just leave the level blank or type a zero for the row in the "Component Prices" column.

This is not actually a "prohibition", because we are not telling the designer to prohibit one level of one attribute from appearing with one level of a different attribute.  Specifying that a certain level should not have any additional incremental price added to it in the Summed pricing does not change the degree of orthogonality of the experimental design.

Please let me know if you are doing something different from my assumption of the use of Summed Pricing above.
answered Aug 21, 2019 by Bryan Orme Platinum Sawtooth Software, Inc. (175,315 points)
Dear Bryan,

Thank you for sharing your thoughts.

Your assumption is right that I’m using Summed Pricing.

Let me explain the difficulty I’m facing for this problem set. In the study we are having 3 attributes including price. first attribute have 4 levels with component Prices for each level and attribute 2 have 5 levels  and for Price we don’t have common Base price but want to have incremental price from $2.5 to $15.0 which interval of $2.5 i.e. $2.5,$ 5.0,$7.5......,$15.0 based on component Prices. For Pricing whenever Level 5 of attribute 2 is displayed we want to only show Component price as per assigned for Attribute 1 and when other levels of Attribute 2 are shown Price should always need to increase between $2.5 to $15.0 from component prices.

For example:

Attribute 1
Level 1 - $50
Level 2 - $100
Level 3 - $150
Level 4 - $100

Attribute 2
Level 1
Level 2
Level 3
Level 4
Level 5

Attribute 3 (Price)  No Base Price but incremental price from $2.5 to $15.0 which interval of $2.5 i.e. $2.5,$ 5.0,$7.5......,$15.0

Let say in first screen 3 combination are shown

Attribute 1 with Level 1 & Attribute 2 with Level 5 then Price should always be $50 ($50.0 + $0.0)
Attribute 1 with Level 2 & Attribute 2 with Level 2 then Price should always be $110 ($100.0 + $10.0)
Attribute 1 with Level 3 & Attribute 2 with Level 4 then Price should always be $52.5 ($50.0 + $2.5)

Please help and suggest for this example in ACBC.
A few thoughts:

1) ACBC isn't usually recommended for studies with fewer than about 4 or 5 attributes.  With 3 attributes, ACBC seems too complex for what can be done more straightforward with CBC.  However, CBC doesn't automatically support "summed pricing", so that indeed may indicate the use of ACBC.

2) Summed pricing is only used to create exact prices for certain product profiles in the BYO (build your own) section of the survey.  After that, the summed prices are used, but then are necessarily modified by a random shock factor of +30% to -30%.  That random shock is needed so that price sensitivity can be estimated separate from the other attributes in the study.

3) If you need to accomplish something specific with the summed pricing aspects of the design that is more tricky than simply summing across the attribute prices independently, then you can implement extra rules for what happens to the component prices when certain attribute levels come together.  You do this using the "Price Adjustments" button on the "Pricing" tab within the ACBC Exercise settings.