To create a implicit study you first need to create a Sentient Prime account. Follow the steps below to get started.
To start building a new study, you can either click the “Create New Study” button on both the homepage or Studies page.
The basic workflow to adding implicit to your research is (1) design and build the study, (2) test & edit the study, (3) integrate the survey links, (4) launch the study, and (5) analyze the results.
Choose a Study Type
On the initial design page, select the study type. Click on ‘Proceed with Study Type’ to enter your project details. If you do not see an exact match for your needs, select ‘Advanced’ and then ‘Custom’ study. If you are a Sentient Prime power user, you can skip this work flow by clicking ‘Study Settings’.
On the main Design screen, start by selecting a country and a language for your study. If your project requires the same study fielded in multiple countries, you will be able to duplicate this study from the ‘Studies’ page and then change the country/language as well as other items requiring translations. Each country or language is treated as a separate study since it will integrate with standalone survey instruments or sample providers.
Next input how many brands, concepts, messages, products, packages etc you would like to evaluate. You will be able to upload these assets on the next page. As you change this amount, notice how it impacts duration on the study length gauge.
In the next section you will be able to select you’re the measurements you want to include. In this section you have a total of 6 option comprised of 3 metrics and 2 study types. The three metrics are emotional appeal (which reveals positive / negative emotional valance), self-identification (which reveals self-relevance or familiarity), and discrete emotions / attributes (which reveals associative strengths). Learn more about these three metrics on the Notifications page on the platform. The two study types are implicit (isolating subconscious processing) or fast explicit (rapid conscious processing). Both study types provide unique and impactful insights. Choose the technique that best fits the research objective and experimental design.
Give your study a title and a brief description so you will be able to quickly find the study.
Before clicking ‘Build Study’, take note of the estimated length, margin of error, and cost for the study just designed. If the study is longer than your budgeted length, you can try any of the following: (1) reduce number of items being evaluated, (2) remove an implicit and/or explicit exercise, or (3) reduce the number of discrete emotions / attributes. If none of these are an option, then progress with building the study and submit a request to our support team to consult on an incomplete block design by expanding the sample.
Margin of Error
Based on over a decade of executing implicit studies, our database norms indicate that an online sample of 200 participants will produce enough statistical power to produce significant effects. This means that any segment requiring statistical precision should have at least these many participants. If your minimum base cell size is different than 200, enter that amount in the ‘Participant Completes Quota’ box. If your base size is less than 200, you will be able to maintain statistical power by increasing the number of repeated measures within the advanced settings. Click ‘Build Study’ to proceed.
You now need to select the images (brands, products, etc) based upon the number of items inputted in the Design phase. You can use Google image search to find ready-to-use images or click ‘Upload’ to import images from your device. If using Google image search, be sure to select the image from the results pane and then drag it into box below. If importing from your device, make sure that the images have a length or width that is at least 400 pixels in order to ensure sufficient quality within the test. Before adding a new image, make sure that the correct slot is in focus – represented by a blue outline before adding. If you want respondents to be exposed to a video prior to testing, upload the video on this page.
To add an emotion or semantic attribute, type the word in the input box provided. If you select one of the Sentient 26 taxonomy of universal emotions, you might be prompted to clarify the contextual meaning (Excitement vs Boredom or Excitement vs Disgust for example). With Sentient 26 emotions, you will be able to access normative data based upon 10 years of testing. If you type in an attribute, you will see some related words from a Princeton Wordnet database. Select any of these options to review it the resulting synonyms and antonyms are relevant to the research and also a common vernacular. If not, use the 3 dots to delete and then enter the word again into the input bar. If you do not select any options and simply hit ‘Enter’, then that word will appear and you will be prompted input 1 synonym and 2 antonyms. When selecting terms for use in implicit tests, ask yourself: (1) is it relevant to the research objective?, (2) is it colloquial and vernacular (that is, will the word be familiar to the survey participant)?, is it unambiguous (does not have multiple meanings), and is it brief?
Clicking on ‘Test Design’ will allow you to view the participant experience of the study. Closing the browser will bring you back to the control panel to make any edits you want to make.
When you click ‘Build Study’ you will see estimate for study length, margin of error, and cost based on any changes you made to the design during the build phase. Click ‘Proceed’ to accept these estimates and continue.
Congratulations! You just created a key to unlock human behavior! From this screen, you have the ability to set the study to ‘Test’ mode in order to collect up to 20 responses for free. You can also use this balance to test integration with a traditional explicit survey instrument, such as Qualtrics, Confirmit, Decipher, etc. To do this, copy the ‘Participant URL’ and send it to the survey programmers with instructions to replace the testRespondent string with each participant’s unique survey ID s
If you really like looking under the hood, click on Advanced Settings to fine tune your study. We’ve auto-optimized every study to ensure reliability in the results yet sometimes even we need to make some edits in order to execute the perfect experimental design. The first step you will need to make is to turn off auto-optimize. After surviving all of our warnings, you will be able to change the order of the exercises, edit any exercise build, or change the number of repeated measures. You can even add an exercise from this screen.
You are ready to field the study. If you do not have any studies and/or completes in your account, you will need to purchase the study before it is entered into launch mode. You will be able to view the progress of fieldwork from the study dashboard by clicking on your study from the ‘Studies’ menu.