Persona Transparency: Analyzing the Impact of Explanations on Perceptions of Data-Driven Personas

Figures & data

Table 1. Research hypotheses and rationales for each

	Hypothesis	Rationale
H01	Increased transparency increases the credibility of data-driven personas.	Higher transparency results in higher credibility, because the explanations reassure the participants that the information can be trusted.
H02	Increased transparency increases the completeness of data-driven personas.	Showing the explanations results in more “rounded” personas that are perceived to be complete by users in terms of information content.
H03	Increased transparency increases the clarity of data-driven personas.	Information in the persona profiles is easier to understand when explanations are given, thus, increasing perceived clarity.
H04	Increased transparency decreases the empathy of data-driven personas.	Transparent personas are perceived as less human-like, because the explanations contain a degree of technicality that takes away the impression of the persona being a real person.
H05	Increased transparency increases the usefulness of data-driven personas	More transparency results in participants wanting to know more about the persona, because the explanations add informativeness of the persona profile.

Figure 1. Transparent persona “Eva”. The participants were provided a full-sized image that shows each section of the persona profile and the accompanying explanations clearly. Another persona, “Marcus”, was created that only differed by demographic attributes and picture. The explanations were detailed considering the space limitations of the persona profile

Table 2. Explanations provided to participants for the attributes of the persona profiles

Section	Explanation provided
Name	Persona’s name is chosen by retrieving common names from a popular online social network of people with a given age, gender, and country. Tools we use: Python, Pandas, Database
Picture	Persona’s picture is chosen from pictures downloaded from online photobanks, tagged for age, gender, country, and ethnicity. Tools we use: Python, Online photobanks, Database
Demographic information	Persona’s demographic information (age, gender, country) is retrieved from aggregated YouTube viewer statistics of this channel’s videos. Tools we use: Python, YouTube API
Industry	Industry is the most common industry of people on Facebook that correspond to this persona’s age, gender, country, language, and topics of interest. Tools we use: Python, Facebook Marketing API, Database
Education Level	Education level is the most common education level of people on Facebook that correspond to this persona’s age, gender, country, language, and topics of interest. Tools we use: Python, Facebook Marketing API, Database
Marital Status	Marital status is the most common marital status of people on Facebook that correspond to this persona’s age, gender, country, language, and topics of interest. Tools we use: Python, Facebook Marketing API, Database
Topics of Interest	Topics of interest are retrieved by classifying the content to descriptive categories and choosing the most corresponding ones for this persona. Tools we use: Python, Pandas, Scikit-learn (Latent Dirichlet Allocation), supervised machine learning, Database
Most Viewed Contents	Most viewed contents are retrieved from the aggregated view counts of YouTube videos and are chosen to describe the taste of this persona. Tools we use: Python, Database, YouTube API
Quotes	Persona’s quotes are retrieved from the comments of most viewed videos of this persona. Tools we use: Python, Database, YouTube API
Audience Size	Audience size is calculated by searching the number of people on Facebook with similar attributes to this persona, including age, gender, country, language, and topics of interest. Tools we use: Python, Facebook Marketing API, Database

Figure 2. Matrix decomposition carried out using NMF. Matrix V is decomposed into W and H. g denotes demographic groups in the dataset, c denotes content (e.g., videos), and p is the number of latent interaction patterns that are used to create the personas

Figure 3. Flow 1. Two personas were shown to participants, such that the participants were randomly assigned to counter-balanced flows (Flow 1: Showing Male persona first, then Female persona; Flow 2: vice versa). Either of the personas always had explanations enabled

Table 3. Constructs and items. These are based on prior work reported in Salminen et al. (2018) that also addressed scale reliability and validity. α = Cronbach’s alpha

Construct	α	Definition	Items
Credibility	0.90	Measures how realistic the persona appears.	The persona seems like a real person. I have met people like this persona. The picture of the persona looks authentic. The persona seems to have a personality.
Completeness	0.93	Measures how well the persona profile captures essential information about the users.	The persona profile is detailed enough to make decisions about the customers it describes. The persona profile seems complete. The persona profile provides enough information to understand the people it describes. The persona profile is not missing vital information.
Clarity	0.83	Measures how clearly the information is presented in the persona profile.	The information about the persona is well presented. The text in the persona profile is clear enough to read. The information in the persona profile is easy to understand. The persona is memorable.
Empathy	0.94	Measures how well the participant relates to the persona.	I feel like I understand this persona. I feel strong ties to this persona. I can imagine a day in the life of this persona.
Usefulness	0.93	Measures how useful the persona is perceived in the given use case.	I would make use of this persona in my task of creating a YouTube video. I would like to know more about this persona. I can imagine ways to make use of the persona information in my task of creating the YouTube video. This persona would improve my ability to make decisions about the customers it describes.

Table 4. Descriptive information about the participants

Age
Min	Max		Mean
23	50		33.5
Education
Undergraduate	Graduate		Doctoral
275 (66.7%)	119 (28.9%)		18 (4.4%)
Nationality
UK	USA	Canada	Australia
260 (63.1%)	127 (30.8%)	20 (4.8%)	5 (1.2%)
Persona experience
No previous experience		Previous experience
284 (68.9%)		128 (31.1%)

Table 5. Univariate tests for within-subjects effects of nontransparent and transparent persona profiles (df(error) = 1 (439))

Independent Variable	Dependent Variable	F	η^2p	p-value
Transparency	Completeness	13.447	0.030	< 0.001
	Usefulness	0.018	< 0.001	0.894
	Credibility	27.289	0.059	< 0.001
	Clarity	12.009	0.027	0.001
	Empathy	0.012	< 0.001	0.914

Participant Gender is included in the model but not reported due to lack of significant effects

Table 6. Univariate tests for between-subjects effects of nontransparent and transparent persona profiles controlling for the gender of the persona (df(error) = 1 (439))

Independent Variable	Dependent Variable	F	η^2p	p-value
Transparency-Gender	Completeness	9.588	0.021	0.002
	Usefulness	5.252	0.012	0.022
	Credibility	0.191	< 0.001	0.662
	Clarity	4.016	0.009	0.046
	Empathy	4.704	0.011	0.031

Participant Gender is included in the model but not reported due to lack of significant effects.

Table 7. Univariate tests for within-subjects effects of nontransparent and transparent persona profiles (df(error) = 1 (439))

Independent Variable	Dependent Variable	F	η^2p	p-value
Transparency x Transparency-Gender	Completeness	4.401	0.010	0.036
	Usefulness	3.074	0.007	0.080
	Credibility	14.765	0.033	< 0.001
	Clarity	0.684	0.002	0.409
	Empathy	5.082	0.011	0.025

Participant Gender is included in the model but not reported due to lack of significant effects.

Table 8. Hypothesis results. Positive support is denoted with (✓) and lack of support with (-)

	Hypothesis	Result	Brief interpretation
H01	Increased transparency increases the credibility of data-driven personas.	-	Added transparency may raise further questions about the trustworthiness of the information.
H02	Increased transparency increases the completeness of data-driven personas.	✓	Explanations are considered as additional information that enhances the completeness of the persona.
H03	Increased transparency increases the clarity of data-driven personas.	✓	Persona information is easier to understand when explanations are provided.
H04	Increased transparency decreases empathy of data-driven personas.	-	Added transparency does not make the personas less “human-like”.
H05	Increased transparency increases the usefulness of data-driven personas	-	Transparency does not directly improve or worsen the usefulness of the persona.

Figure 4. Eve Persona Profile without explanations. The use of more detailed explanations in the profile of Eve had a positive effect on Completeness and Empathy of the female persona

Figure 5. The racial difference between the personas may be a confounding factor in the analysis of the persona’s gender and effect on perceptions, with Marcus being black and Eva being Asian. This is an area for future research

Salminen, J., Kwak, H., Santos, J. M., Jung, S.-G., An, J., & Jansen, B. J. (2018). Persona perception scale: Developing and validating an instrument for human-like representations of data. CHI’18 Extended Abstracts: CHI Conference on Human Factors in Computing Systems Extended Abstracts Proceedings, Montréal, Canada.

 Google Scholar