Visitor Studies

ISSN: 1064-5578 (Print) 1934-7715 (Online) Journal homepage: www.tandfonline.com/journals/uvst20

Reasoning about Objects in a Natural History

Museum: The Effect of Complexity of Questions on
Object Labels

Anne M. Land-Zandstra, Kelly Hoefakker & Welmoet Damsma

To cite this article: Anne M. Land-Zandstra, Kelly Hoefakker & Welmoet Damsma (2020)
Reasoning about Objects in a Natural History Museum: The Effect of Complexity of Questions
on Object Labels, Visitor Studies, 23:2, 218-236, DOI: 10.1080/10645578.2020.1781485

To link to this article: https://doi.org/10.1080/10645578.2020.1781485

Published online: 02 Jul 2020.

Submit your article to this journal

Article views: 575

View related articles

View Crossmark data

Citing articles: 5 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uvst20
VISITOR STUDIES
2020, VOL. 23, NO. 2, 218–236
https://doi.org/10.1080/10645578.2020.1781485

Reasoning about Objects in a Natural History Museum:

The Effect of Complexity of Questions on Object Labels
Anne M. Land-Zandstraa , Kelly Hoefakkera, and Welmoet Damsmab
a
Department of Science Communication & Society, Leiden University, The Netherlands;
b
Education Department, Naturalis Biodiversity Center, Leiden, The Netherlands

ABSTRACT
In natural history museums, a large part of the educational mission is
to facilitate family learning with and about objects. Questions on
object labels can play a role in this learning process. In the current
study, we investigated the effect that different types of questions on
object labels can have on the reasoning conversations among family
members. We audio-recorded family conversations at an exhibit that
included a fossilized dinosaur egg and a text label containing a ques-
tion. Sixty-six families participated in three conditions differing in the
level of complexity of the question on the label. We found that a
question of moderate complexity facilitated the longest conversa-
tions, with the largest number of complex inferences, compared to
the simpler and the more complex question. When reasoning, families
most often used evidence from prior knowledge and other parts of
the exhibition, but did not often relate to personal experiences.
During the longer and more complex conversations, parents took up
a larger role in the conversation, facilitating their family’s reasoning
process. We suggest that open-ended, moderately complex questions
on labels can facilitate family reasoning conversations. Providing
enough context within the surrounding exhibits and connecting to
prior knowledge may help the reasoning process.

Introduction
In a time in which internet and virtual reality are providing information and experien-
ces at everyone’s fingertips, museums in general and science museums in particular still
have a unique position because they provide real objects (Dillon et al., 2016). Over the
years, science museums have evolved from cabinets of wonder, showing many different
objects without much context, to educational institutions that want to provide learning
experiences that teach, amaze, and change visitors (Friedman, 2007; Packer &
Ballantyne, 2002). For these educational goals, just showing objects is not sufficient. The
learning process of visitors needs to be facilitated by supporting materials, such as
object labels, images, video, audio, or interactives. In addition, social interaction sup-
ports the learning process as well (Vygotsky, 1978). This social learning process happens
through conversations among visitors or with museum staff. Conversations are both an
instrument for and a demonstration of learning (Hohenstein & Tran, 2007). Visitors

CONTACT Susan M. Land sland@psu.edu 317 Keller Building, Penn State University, University Park, PA
16802, USA.
ß 2020 Visitor Studies Association
 VISITOR STUDIES 219

may learn more when they talk with each other about what they see or experience
(Allen, 2002; Crowley et al., 2001). Simultaneously, the content of conversations can be
an indication of what is learned in museums (Hohenstein & Moussouri, 2018). In fam-
ily learning conversations, parents play a key role in the learning process of their chil-
dren (Gleason & Schauble, 1999; Rogoff, 2003; Zimmerman et al., 2008). Exhibit labels
could provide a means to facilitate this learning process of parents and children.
Previous studies have looked at exhibit labels and family conversations related to inter-
active exhibits (Allen, 2002; Horn et al., 2016). However, few studies have focused on
how text labels may support the interaction and conversations around authentic objects
(e.g. Hohenstein & Tran, 2007; Serrell, 2015). In the current study we examined in what
way questions on labels with a natural history object could facilitate reasoning in con-
versations between parents and children.

Theoretical framework
Science learning in museums
Learning in (science) museums is often studied from a social constructivism perspective
(Hein, 1998; Hohenstein & Moussouri, 2018), stemming from the work of Vygotsky
(1978). Such studies look at learning in museums as a social process where knowledge
and understanding is co-constructed among group members (Hohenstein & Moussouri,
2018). This makes sense because in many science museums, people visit within (family)
groups and interact with each other and with the museum objects and exhibits during
their visit, while learning together. For example, Horn et al. (2016) found that social
interaction during exhibit engagement had a positive impact on learning about evolu-
tion. Falk and Dierking (2000) propose in their Contextual Model of Learning that
learning in museums is complex and is impacted by several contexts of the visitor and
the museum itself. Learning is a combination of a process that happens within a person
and input from outside (e.g. other visitors, the exhibits, events happening, conversa-
tions, other experiences from the past).
In addition, within museum education, learning is often considered in a broader
sense than just the acquisition of new information (Kisiel et al., 2012; National Research
Council, 2009; Rowe, 2002). Learning in museums is often considered to include more
affective outcomes such as creativity and interest as well as acquiring skills such as
being able to ask researchable questions or reasoning with evidence (National Research
Council, 2009). In this study, we focus on reasoning, since it is an important aspect of
scientific thinking and scientific literacy (Kisiel et al., 2012).

Reasoning
Scientific reasoning is a significant aspect of science literacy, especially reasoning with
evidence. It includes asking questions, forming hypotheses, collecting evidence, making
inferences, and revising theories (Crowley et al., 2001; Kisiel et al., 2012). Kisiel et al.
(2012) define reasoning as the “gathering and use of evidence for thinking and commu-
nicating about thinking”. Zimmerman (2000) mentioned in her review of the develop-
ment of scientific reasoning skills that participants used their prior experiences and
220 A. M. LAND-ZANDSTRA ET AL.

conceptual knowledge if a scientific reasoning task contained a question or a problem.

The knowledge of and the comparison with a prior situation was then used as evidence
in reasoning. Gillies and Khan (2009) argued that teaching children to ask and answer
questions is important to engage children in argumentation, problem-solving and learn-
ing in general.
Previous studies have looked at scientific reasoning in informal science education
(Allen, 2002; Ash, 2003; Gutwill & Allen, 2009; Kisiel et al., 2012). For example, Gutwill
and Allen (2009) looked at how a specific intervention can increase the reasoning
behavior of visitors. Other studies such as the ones by Ash (2003) and Allen (2002),
focused on what reasoning looked like in family conversations.

Family conversations in museums

Within a social constructivism perspective of learning, conversations play an important
role. Conversations can both be part of the process of learning as well as the outcome
of learning (Leinhardt et al., 2002). As visitors interact with each other, their conversa-
tions are part of the learning process where they try to make sense of what they see. At
the same time, listening to their conversations may be a way to gauge what they have
learned during their visit. As such, visitor conversations have become part of the study
of learning in museums.
Allen (2002) and Crowley et al. (2001) argued that scientific reasoning and science
literacy development will be supported by family conversations in informal learning set-
tings. Since parents know their children, they can provide links with previous know-
ledge and experience.
Parents can play different roles in family learning conversations. Firstly, parents can
take the role of explainer or idea-suggester (Gleason & Schauble, 1999; Rogoff, 2003).
This can support the short-term learning of their child (Zimmerman et al., 2008).
Rogoff (2003) called this process guided participation, which means that children can
participate in science guided by their parents. Secondly, Bell et al. (2006) suggested that
parents can act as bridges from children’s prior experiences to their expertise develop-
ment in science, technology, engineering and mathematics (STEM). The facilitation by
parents includes the ability to remind their children of prior experiences. These familiar
experiences can be used to explain or define new observations (McClain &
Zimmerman, 2014). Storytelling, by both children and parents, can also help to connect
to and share prior experiences and knowledge in the learning process (Goodwin, 2007).
Thirdly, focusing attention on important aspects and details of an object by parents sup-
ports children’s understanding (Ornstein et al., 2004). Fourthly, similarity comparison
of the unfamiliar object to familiar information can also support meaning-making of
that object. The cognitive tool of similarity comparison may increase children’s concep-
tual development (Valle & Callanan, 2006). Finally, Ash (2003) showed that parents use
a variety of inquiry skills, such as observing, questioning, hypothesizing, explaining and
interpreting, when they talk with their children. Since children may take over the behav-
ior of their parents, they can learn many scientific skills by the conversations with
their parents.
 VISITOR STUDIES 221

With regard to scientific reasoning in family conversations, Crowley et al. (2001)

showed that children’s exploration of evidence was observed longer, broader and more
focused on comparisons when they engaged with an exhibit with their parents than
without their parents. The researchers saw that parents talked about evidence identifica-
tion, generation and interpretation to their children, which may help children to select
relevant evidence. In some cases, parents took the role of explainer. In addition, parents
helped children to build explanations about science during informal science activities.
All of these strategies invite children to engage in scientific reasoning and to develop
the habits of scientific literacy (Crowley et al., 2001).

Object labels
In this study, we look at reasoning in family conversations regarding an authentic object
in a natural history museum. Although traditionally, museum objects were considered
to “speak for themselves”, that assumption may not be true (Evans et al., 2002). Visitors
may look beyond just the appearance of an object and also consider its history and
story (Van Gerven et al., 2018). As Evans et al. (2002) point out, the context in which
objects are presented may impact the “voice that is being heard” by the visitor. In add-
ition, context, including object labels, can influence the interaction with museum
objects. Information on object labels, can support family conversations. Rand (2010)
found that parents read out loud the labels to their children. This could be a start of a
conversation about the object, because parents focus the attention of their children to
the information on the label (the third category mentioned above). Moreover, Atkins
et al. (2009) suggested that parents explained more to their children in the presence of
object labels than without. Visitors utilized the authority of the science museum by
reading the labels to confirm their emerging scientific claims during a conversation
(McClain & Zimmerman, 2014; Zimmerman et al., 2010). Hohenstein and Tran (2007)
argued that guiding questions as part of the label text at object-based exhibits might
stimulate inquiry and reflection in conversations. They found that exhibit labels with
simplified text and one question, and exhibit labels with a guided question led to a dia-
logue with more open-ended questions. In addition, labels with a guided question led
more often to explanations than labels with a simplified text (Hohenstein & Tran,
2007). However, their results were ambiguous between different exhibits.

Questions on object labels

Earlier studies suggested that questions on object labels in museums can stimulate meaningful
conversations (Gutwill, 2006; Hohenstein & Tran, 2007). Haden et al. (2014) argued that ques-
tion-asking and answering is important to stimulate understanding and learning of a scientific
problem. However, the formulation of questions on the label might influence the conversa-
tion’s content. Fivush et al. (2006) focused on open-ended questions and concluded that ask-
ing open-ended questions invited children to actively participate in the conversation. Open-
ended questions often start with Who, What, Where or How (Fivush et al., 2006; Haden,
2010; Jant et al., 2014). These open-ended questions invite people to go beyond listing facts or
naming objects, to conversations involving making meaning of what they observe. Haden
222 A. M. LAND-ZANDSTRA ET AL.

et al. (2014) concluded that open-ended questions were more cognitively challenging than
closed-ended questions, because they invited children to generate their own explanations.
Rand (2010) described five types of questions that fostered interactivity and family
learning: 1) attention-focusing questions and measuring or counting questions, 2) com-
parison questions, 3) action questions, 4) problem-posing questions, 5) reasoning ques-
tions. As discussed above, in our study we selected reasoning questions to investigate if
such questions would indeed ignite reasoning conversations.

Current study
In order to examine the effect of reasoning questions on object labels on family conver-
sations, the current study analyzed spontaneous family conversations in a Dutch natural
history museum. Our research question was formulated as follows: What is the effect of
open-ended questions with different levels of complexity on the reasoning within family
conversations in a natural history museum?
We addressed several subquestions:

What is the effect of questions with different levels of complexity on the length
of the conversations?

What type of talk do families adopt (i.e. simple or complex inferences)?

What type of evidence do families use (e.g. focus on comparisons or broaden-
ing content)?

Methods
Research setting
Naturalis Biodiversity Center in Leiden, the Netherlands, is a natural history museum
that aims at encouraging curiosity and scientific literacy related to the natural world.
Our study was performed during the exhibition T.rex in Town (2016-2017), centering
around the display of a real T.rex skeleton. In order to investigate the impact of labels
on family conversations, we selected a fossil of a dinosaur egg as the object of study,
because it was an object that linked to the topic of the exhibition, and could speak to
children’s imagination. Furthermore, the shape and size could play a role in the conver-
sations, giving starting points for reasoning. The object was placed by itself on a table
in the same hall as the real T.rex skeleton (see Figure 1). An audio recorder was set up
behind the table to record the conversations. One of the researchers was present in the
exhibition hall to unobtrusively observe the families.

Object labels
We designed object labels with only one question to decrease distraction. We did not
provide the answer in order to be able to focus on the impact of the question alone. In
addition, even scientists are not sure about the answers to these questions. The text on
the labels was limited, so that families did not stop reading because the text was too
long. The questions for each condition were:
 VISITOR STUDIES 223

Figure 1. The dinosaur egg that was used in the experiment.

Condition A. ‘What does this look like?’

Condition B. ‘This is a REAL dinosaur egg. How big do you think the mother of
this dinosaur egg was?’

Condition C. ‘This is a REAL dinosaur egg. What would the breeding place of a
dinosaur have looked like?’

The questions differed in the level of complexity and the amount of reasoning that
was needed to answer it. The first condition served as a control (no reasoning needed);
the second condition needed a little bit of reasoning (comparing the T.rex with familiar
animals and egg sizes); and the third condition needed more reasoning (comparing with
other animals and reasoning about living environments) and possibly also more back-
ground knowledge about these living environments. The message that the object is a
real object was included in the last two conditions to prevent the conversations from
only focusing on the realness of the objects rather than on the question itself.

Participants
Since families are the main visitor group of the museum, we targeted families as partici-
pants for this study. In order to not introduce any experiment bias, we randomly
recruited 118 families at the entrance of the museum. Families were defined as at least
one adult with at least one child in the age of primary school (4-12 years). They could
include older children as well. Only Dutch-speaking families were included in the study.
Whenever a family that seemed to include children of the right age group approached
the entrance of the museum they were asked to participate in the study. Families were
asked to participate in the research study and were told that they would be recorded at
some point in the exhibition, but no other information was given to make sure that
they would not act differently because they were part of a study. The exhibit under
study was located around half way through the exhibition. After one of the adults gave
written consent, participating families were identified by a sticker on their clothing so
that the researcher could observe them during the conversations at the research set up
and could identify which audio recordings to include in the study. Not all families
224 A. M. LAND-ZANDSTRA ET AL.

Table 1. Demographics and group size across conditions.

A B C Total
% female adult 53.3% 59.4% 64.7% 59.4%
% female child 42.5% 38.6% 34.8% 38.5%
Mean age children 8.5 8.6 8.6 8.6
Mean group size 3.0 3.5 3.6 3.4
Note. Condition A: What does it look like? Condition B: How big was the mother? Condition C: What would the breed-
ing ground look like?

interacted with the exhibit under study, so we were able to record conversations of 66
families, evenly spread out over the three research conditions (23 in condition A, 21 in
condition B, 22 in condition C).
On average families consisted of 3.4 group members (range 2–8). In all three condi-
tions, group sizes were similar (3.0, 3.5, 3.6 respectively). Of all participating children,
62% (82 children) were boys and 38% (50 children) were girls. Of the parents who gave
permission for their family, 35% (24 adults) were male and 65% (44 adults) were female.
Some families included more than one adult, but we only collected information from
the consenting adult. The age range of the children was 4–15 years old, with an average
of 8.6 years old. Table 1 shows that these family characteristics were similar across the
three conditions. Because we recorded spontaneously occurring conversations, not all
families were complete during the recorded conversation. Generally, recorded conversa-
tions included two to four members of the family. The majority of conversations
included at least one adult and one child.

Data analysis
The recorded conversations were transcribed and coded. We coded for length of con-
versation (unit of analysis: conversation, starting when the first person of a family
started to talk and ending when the family walked away from the exhibit; measured in
seconds), portion of the conversation by parents versus children (unit of analysis: con-
versation; measured in number of words), and content of the conversation (unit of ana-
lysis: fragments). Two common approaches for analyzing content of conversations are
content analysis and discourse analysis. Herrera and Braumoeller (2004) describe that
although the two approaches come from different philosophical backgrounds, there are
many similarities and overlaps. Because we were interested in the content of the conver-
sation (evidence and content of reasoning) and not so much the way language was
being used, and because we were using a mostly top-down coding scheme, we decided
to implement a content analysis approach. However, we did incorporate some charac-
teristics usually assigned to discourse analysis (studying conversations and not written
texts, focusing on the meaning of text and not just counting words).
For the content analysis we divided the transcript up in fragments that contained one
statement or one inference. Sometimes an inference could be an exchange among sev-
eral members of the conversation developing one argument. To analyze the content of
the conversation we adapted the codebook of Allen (2002). She developed a framework
to investigate learning conversations in museums. The framework consists of five types
of talk. Perceptual talk are all statements meant to draw attention to an element of the
exhibition, e.g. pointing at it, naming it, or reading the text label aloud. Conceptual talk
 VISITOR STUDIES 225

refers to cognitive interpretations such as reasoning, inferring, predicting. Two types of infer-
ences were distinguished: simple and complex inferences. Simple inferences are “single interpret-
ative statements of interpretation”, often based on an observation or prior knowledge. Examples
of simple inferences are: “It’s an egg”, “It looks like a rock”. Complex inferences go beyond sim-
ple interpretations and contain a hypothesis, generalization or comparison. Examples of complex
inferences are: “I think there is a real dino inside” or “But in those times there were different
eggs”. Connecting talk is talk that connects the exhibit to some other knowledge or experience,
inside or outside of the museum. Connecting talk could connect to another exhibit (inter-exhibit
connecting talk; IECT), to visitors’ personal lives (life connecting talk; LCT) or to prior know-
ledge (knowledge connecting talk; KCT). Strategic talk referred to how to interact with the
exhibit (adapted to mean how to interact with an object-based exhibit instead of an interactive
exhibit). Affective talk were all utterances that conveyed some sort of emotion. Because we were
interested in reasoning (gathering and use of evidence for thinking) in family conversations, we
defined simple and complex inferences as reasoning and used connecting talk to identify which
evidence families were using within their reasoning conversation. Therefore we double coded
each inference with a label for the type of inference (single or complex) as well as the type of
connection (inter-exhibit, life, or knowledge). Appendix A contains the codebook.
Intercoder reliability was tested by having a second coder code 29% of the data. Cohen’s
kappa was determined at 0.72 which indicates a satisfactory reliability. The conversations
were transcribed, coded and analyzed in Dutch. We translated relevant quotes to English for
this article. In order to investigate differences among the three conditions we performed
some statistical analyses where appropriate. The number of conversations in each condition
was not large enough to perform a statistical comparison of average holding time in
each condition (as determined by a power analysis), but we did perform chi square tests to
determine significant differences in number of words spoken by parents versus children,
and number of simple versus complex inferences. We inspected standardized residuals to
determine which comparisons were significantly different from the expected values.

Results
General analysis
Although they were not specifically prompted to read the label, all participating families
that had a conversation at the exhibit under study referred to the information or question
on the label. In condition A (‘What does this look like?’), 20 out of 23 families (87%) read
the label out loud and 22 families (96%) tried to answer the question. In condition B (‘This
is a REAL dinosaur egg. How big do you think the mother of this egg was?’), 18 out of 21
families (86%) read the label out loud, but only 15 families (71%) actually discussed the
question. In condition C (‘This is a REAL dinosaur egg. What would the breeding place of
a dinosaur have looked like?’) only 9 out of 22 families (41%) read the label out loud and
only 5 families (23%) tried to answer the question.
Figure 2 shows the length of the family conversations for each condition. Overall,
most conversations stopped after 60 seconds. In condition A (What does this look like?)
the average holding time was 45 seconds (range: 9-127 s). The average holding time in
condition B (How big was the mother?) was 52 seconds (range: 19-148 s). In condition
C (What does the breeding place look like?) the average holding time was 45 seconds
226 A. M. LAND-ZANDSTRA ET AL.

Figure 2. Boxplots of conversation time (s) for each condition. A: “What does it look like?” (M ¼ 46.7s). B:
“How big was the mother?” (M ¼ 52.7s). C: “What would the breeding ground look like?” (M ¼ 47.0s).

Figure 3. Average number of words uttered by parents and children per conversation for
each condition.

(range: 7-138 s). Since families in condition B had longer conversations, they used more
words on average (M ¼ 117; range: 12-440) than in condition A (M ¼ 68; range 11-249)
and C (M ¼ 64; range 5-175). Because of the number of families in each condition, no
statistical tests were performed to determine if these differences were significant. A Chi
square test on the number of words spoken by children versus parents in each condi-
tion showed that in condition B, children spoke significantly less while parents spoke
more than would be expected based on the entire sample (v2(2,5261)¼24.4, p<.05).
Figure 3 shows the number of words for each condition.

Content analysis
All conversations were coded for the type of talk that was used. Almost all conversa-
tions included some kind of perceptual talk (i.e. directing attention; 97%), conceptual
 VISITOR STUDIES 227

Table 2. Content of conversations.

Percentage of conversations (%) Average number of utterances per conversation
Perceptual talk 97.0 3.1
Conceptual talk 95.5 5.4
Connecting talk 86.4 3.6
Strategic talk 56.1 1.5
Affective talk 12.1 0.2

talk (i.e. cognitive interpretations; 95.5%) and connecting talk (i.e. connection to know-
ledge/experience; 86.4%), see Table 2. Strategic talk (i.e. on how to interact with the
exhibit; 56.1%) and affective talk (i.e. emotions; 12.1%) were less present in the conver-
sations. Another way of looking at the type of talk is by looking at the average number
of utterances per conversation, to see how prevalent each type of talk was throughout
the conversations. This analysis showed that conceptual talk (M ¼ 5.4 utterances) was
the most prevalent, followed by connecting talk (M ¼ 3.6) and perceptual talk (M ¼ 3.1).
Again, strategic talk and affective talk were rare (M ¼ 1.5 and M ¼ 0.2, respectively).
Conceptual and connecting talk were most relevant to our research questions about
reasoning, therefore they will be elaborated on below.

Type of reasoning
Within the conceptual talk we determined the type of reasoning in each conversation
by distinguishing between simple inferences (SI) and complex inferences (CI). An
example of a SI uttered by a child in this study is: “It looks like the brain of a T.Rex”
(A21, c; child from family 21 in condition A), because the child made the assumption
that the object must be part of the T.rex skeleton, but no second reasoning step was
taken. We also considered statements like “brains”, “eyeballs”, “cranium” or “dung” SI
because these statements must be based on recognizing features, shapes, size and color
of something familiar, even if the reasoning was not explicitly part of the conversation.
An example of a complex inference is: “Is there a dinosaur left inside then?” (C12, c).
This question includes a CI, because it was about the concepts related to the object.
Knowledge about eggs (there should be something in it), is used to come to this ques-
tion. This reasoning step made it a CI. Other examples of complex inferences are: “But
do you think it would have such a nest, like a bird, in a tree?” (C3, p) and “It [dino
poop] is not from this one [the T.rex], because it is not that big.” (A15, c)
Figure 4 shows the average number of inferences per conversation for each condition.
The number of simple inferences per conversation was highest in condition A (M ¼ 3,3
SI per conversation) followed by condition B (M ¼ 1,0) and C (M ¼ 0,9). In contrast,
complex inferences were most prevalent in condition B (M ¼ 2,4 CI per conversation)
followed by condition C (M ¼ 0,9) and condition A (M ¼ 0,5). Chi square tests showed
that in condition A, there were more simple inferences and less complex inferences
than could be expected based on the entire sample, while in condition B there were
more complex inferences and less simple inferences (v2(2,196)¼56.3, p<.001).
Condition C did not show a significantly different amount of simple or complex infer-
ences compared to the entire sample.
228 A. M. LAND-ZANDSTRA ET AL.

Figure 4. Average number of simple inferences (SI) and complex inferences (CI) per conversation for
each condition.

Evidence in reasoning
For each inference, we analyzed what type of evidence was used. We distinguished
knowledge connecting talk (KCT), inter-exhibit connecting talk (IECT) and life
connecting talk (LCT). Below is an example of a complex inference with
knowledge connection. The quote suggested that the boy knows that fossils often
will be glued.
Boy: “But it does not hatch anymore”
Mother: “No”
B: “Because it is glued” (B26)
Another conversation shows an example of inter-exhibit connection talk. In this con-
versations in condition A, a mother and a daughter discussed what the object could be.
The girl suggested that the object would be a fracture of the T.rex ribs which they could
see behind them.
Girl: “Yeah I know it! Such a bump.”
Mother: “A bump of which?”
G: “Of uhm. The ribs things or something … ”
M: “Where the ribs are broken do you mean?”
G: “Yes, those bump.”
M: “Don’t you think that is a very big bump if you look at the ribs? Seems a bit too big
to me.”
G: “Yes, but look how big he is!”
M: “Yes, he is very big.” (A40)
Finally, an example of an inference with life connecting talk is shown below where a
grandmother talks about something she had seen somewhere else. Only the statements
where a specific life situation is mentioned were labeled as life-connecting.
 VISITOR STUDIES 229

Figure 5. Average number of connecting talk per conversation in each condition, separated in life
connections, knowledge connections, and inter-exhibit connections.

GM: “Should it be as big as an ostrich? Yes, I saw that in Aruba. But I had the idea that
they were bigger … ” (B11)
Figure 5 shows the number instances of connecting talk within each condition.
Knowledge connection was the most common connecting talk used during this study
(72% of all connections), followed by inter-exhibit connection (24%) and life connec-
tions (5%). Conversations in condition B contained the most instances of connecting
talk (M ¼ 4.8 per conversation), followed by condition A (M ¼ 4.0) and condition
C (M ¼ 2.0).

Discussion
In this study, we investigated the effect of reasoning questions on family conversations
in a natural history museum. In particular, we provided object labels in three different
conditions with varying levels of reasoning and background knowledge needed to
answer the question. We then looked at the reasoning in the family conversations at the
exhibit. In condition A, the label read “What does this look like?”; in condition B it
read: “This is a REAL dinosaur egg. How big do you think the mother of this egg was?;
and in condition C it read: “This is a REAL dinosaur egg. What would the breeding
place of a dinosaur have looked like?”

Conversation characteristics
Many of the families in our study who stopped at the dinosaur egg exhibit and had
conversations about the object read the label out loud, supporting Rand’s (2010) sugges-
tion that parents use labels in the interaction with their children. More than half of the
families in our study actually tried to answer the question. In condition B (about the
mother of the egg), we heard the longest conversations, and parents talked more than
their children compared to the other two conditions. These outcomes suggest that con-
dition A was a question that could be solved with a short answer without a lot of sup-
port from a parent. Condition B was complex enough to trigger parents and children to
230 A. M. LAND-ZANDSTRA ET AL.

engage in longer and more complex conversations. Condition C may have been too
complex to trigger parents to facilitate a conversation, as we discuss below.

Content of conversations
When looking at the type of talk in the family conversations, conceptual, connecting
and perceptual talk were most prevalent, meaning that people talked most about cogni-
tive interpretations, connections to prior knowledge or experiences and directing atten-
tion. Talk about emotional reactions (affective talk) and about interaction with an
exhibit (strategic talk) were rare. It is difficult to compare these percentages to the ori-
ginal study by Allen (2002), because the situations are very different (complete exhib-
ition vs. one exhibit; percentage of stops vs. percentage of utterances or conversations;
interactive exhibits vs. object-based exhibit, making strategic talk less relevant).
However, one clear difference is the low prevalence of affective talk in our study com-
pared to Allen’s, where affective talk was the second most heard talk. In addition, in
Allen’s study connecting talk was less common than in our study. Possibly, the fact that
we specifically invited people to start reasoning about the provided question, resulted in
more talk that connected to prior knowledge compared with the spontaneous conversa-
tions in Allen’s study. Coupled with the fact that we heard most of the people reading
the question and trying to answer it, this suggests that the questions on the label were
able to facilitate family conversations toward connecting to prior knowledge and
experiences.

Reasoning in conversations
With regard to the reasoning that families conducted within their conversations we
found that in condition A most inferences were simple inferences, in condition B most
inferences were complex, and in condition C both types of inferences were low. In con-
dition A, the question triggered comparison of the unfamiliar object to familiar infor-
mation. This type of comparison can support meaning-making of the dinosaur egg
(Valle & Callanan, 2006). The comparisons that families made were all linked to the
physical features and the shape of the object. In condition B, where the label gave away
what the object was, the number of simple inferences was lower than in condition A.
The question in condition B triggered more complex inferences where parents and chil-
dren together reasoned about how big the mother of the dinosaur egg could have been.
These complex inferences often included several consecutive statements, going back and
forth between parents and children.
Although conversations were longer in condition B, the total number of statements
was lower than in condition A, caused by the fact that complex inferences take a longer
time to develop. In contrast to condition A and condition B, in condition C the total
number of inferences made by the families was much lower. We speculate that the
chosen word ‘breeding place’ (in Dutch: “broedplaats”) was too difficult to understand
for children, or was not something that they themselves would wonder about. As Serrell
(2015) points out, questions on labels work best when the questions are similar to what
visitors ask spontaneously. Also, parents may have had more trouble imagining a
 VISITOR STUDIES 231

dinosaur nest than imagining the mother of an egg. They also needed to have some
prior knowledge about the living environments of dinosaurs, which may have been lack-
ing among the participating adults. As Ash (2003) showed, parents often take the lead
in reasoning and experimenting with their children. We expect that parents may be
hesitant to start a reasoning conversation with their kids if they don’t know beforehand
to which answer they are headed or if they feel they lack background knowledge.
Parents may need some help in starting conversations with their children when they do
not know the answer themselves. Similarly, Shine and Acosta (2000) found that parents
had a more educational goal in mind (teaching their children something), instead of
engaging in pretend play with their children. The authors suggest that parents need spe-
cific guidance to engage in open-ended play with their children. The study of Jant et al.
(2014) showed that parents who had been given cues on how to have conversations
with their children about objects talked with their children in more elaborate ways.

Connecting talk
We also looked at what evidence families used in their reasoning, i.e. to what type of
information they connected. In all three conditions, most connections were made with
prior knowledge (knowledge connecting talk, KCT), followed by connections to other
exhibits (inter-exhibit connecting talk, IECT) and only a few connections with experien-
ces in visitors’ personal lives (life connecting talk, LCT). Connections with prior know-
ledge often included links with familiar objects such as chicken, ostrich, other eggs,
other nests or other types of knowledge that visitors had about dinosaurs and eggs (e.g.
that eggs are often glued together to preserve them, that larger animals often have larger
eggs). These outcomes suggest that it is important to know about prior knowledge and
experiences of the audience, for example through front-end evaluations (Borun, 2002).
When family conversations included IECT, they used mostly the T.rex skeleton stand-
ing next to the object in their reasoning. Some families walked around the skeleton to
find evidence for what the object could be or to compare the features of the object to
the bones of the T.rex. This observation suggests that it is important to provide clues
for reasoning nearby, so that visitors’ can use evidence from surrounding exhibits in
their reasoning conversations. Making sure that an exhibition has a clear overarching
theme or storyline may help (Allen, 2002). Contrary to what other studies suggest
(Allen, 2002; McClain & Zimmerman, 2014), the participants in this study did not
make many connections to their personal lives. Possibly, the type of object, the topic
(an extinct animal) and the type of assignment (invitation to start reasoning) did not
trigger many links to personal experiences.

Limitations
This study has some limitations that may have impacted the validity of the results. One
limitation is the fact that there might be a bias in the type of families that agreed to
participate in this study. In addition, even though we did not give away the exact pur-
pose of the study, they knew they were going to be audiotaped at some point in the
museum. This may have impacted their behavior. However, none of the recorded
232 A. M. LAND-ZANDSTRA ET AL.

families noticed the microphone, and only a few families speculated that the dinosaur
egg exhibit might be the research location. Another limitation were technical issues
such as the crowdedness of the exhibit at some moments and the fact that sometimes
the conversations continued outside of the range of the microphone, resulting in some
missing data. Lastly, in order to be able to perform statistical analyses on holding time
of the object and questions a larger sample per condition should have been included.

Future research
One of the avenues of future research is to perform similar studies with different types
of objects. As discussed, we suspect that visitors may use different types of evidence for
different types of objects (e.g. more connections to personal lives with more familiar
objects). In addition, it would be interesting to find out more about what level of rea-
soning parents and children are able to reach on their own. It would be interesting to
find out why parents are reluctant to start reasoning conversations with their children if
they do not know the answer themselves. Lastly, investigations into the interaction
between parents and children during these conversations are another possible topic for
future research. What role does each of the conversation participants take and who
takes the lead?

Conclusion
In this study we found that open-ended reasoning questions on a label accompanying a
dinosaur fossil was able to facilitate reasoning conversations among families. The type
of complexity of the question had an effect on the length and content of the conversa-
tions. The question with a moderate complexity (“How big do you think the mother of
this dinosaur egg was?”) seemed to trigger longer and more complex reasoning conver-
sations than the simpler question (“What does this look like?”) and the more complex
question (“What would the breeding place of a dinosaur have looked like?”).
Furthermore, during longer reasoning conversations, parents took up the role of facilita-
tor, illustrated by the fact that their share of the conversation is larger than their child-
ren’s. During their reasoning, family members connected mostly to prior knowledge
and other parts of the exhibition and not so much to their personal lives. Our results
suggest that in order to support family reasoning conversations, exhibits should include
open-ended questions of a suitable complexity and should provide enough links or con-
text within the surroundings of the exhibit for people to make connections to prior
knowledge and other elements within the exhibition. Parents could be supported by giv-
ing them hints of where to start their reasoning.

Acknowledgements
We would like to thank Dylan van Gerven and Tim van Wessel for their help with this research
and the manuscript.
 VISITOR STUDIES 233

Disclosure statement
No potential conflict of interest was reported by the authors.

ORCID
Anne M. Land-Zandstra http://orcid.org/0000-0002-7604-9092
Welmoet Damsma http://orcid.org/0000-0002-5290-0663

References
Allen, S. (2002). Looking for learning in visitor talk: A methodological exploration. In G.
Leinhardt, K. Crowley, & K. Knutson (Eds.), Learning conversations in museums (pp.
259–303). Lawrence Erlbaum Associates.
Ash, D. (2003). Dialogic inquiry in life science conversations of family groups in a
museum. Journal of Research in Science Teaching, 40(2), 138–162. https://doi.org/10.1002/
tea.10069
Atkins, L. J., Velez, L., Goudy, D., & Dunbar, K. N. (2009). The unintended effects of interactive objects
and labels in the science museum. Science Education, 93(1), 161–184. https://doi.org/10.1002/sce.20291
Bell, P., Bricker, L. A., Lee, T. R., Reeve, S., & Zimmerman, H. T. (2006). Understanding the cul-
tural foundations of children’s biological knowledge: Insights from everyday cognition research.
In June 26-July 1 (Ed.) [Paper presentation]. International Conference of the Learning
Sciences, Bloomington, IN.
Borun, M. (2002). Object-based learning and family groups. In S. G. Paris (Ed.), Perspectives of
object-centered learning in museums (pp. 245–260). Lawrence Erlbaum Associates.
Crowley, K., Callanan, M. A., Jipson, J. L., Galco, J., Topping, K., & Shrager, J. (2001). Shared sci-
entific thinking in everyday parent-child activity. Science Education, 85(6), 712–732. https://
doi.org/10.1002/sce.1035
Crowley, K., Callanan, M. A., Tenenbaum, H. R., & Allen, E. (2001). Parents explain more often
to boys than to girls during shared scientific thinking. Psychological Science, 12(3), 258–261.
https://doi.org/10.1111/1467-9280.00347
Dillon, J., DeWitt, J., Pegram, E., Irwin, B., Crowley, K., Haydon, R., King. H., Knutson, K.,
Veall, D., & Xanthoudaki, M. (2016). A learning research agenda for natural history institutions.
Natural History Museum.
Evans, E. M., Mull, M. S., & Poling, D. A. (2002). The authentic object? A child’s-eye view. In
S. G. Paris (Ed.), Perspectives on object-centered learning in museums (pp. 55–77). Routledge.
Falk, J. H., & Dierking, L. D. (2000). Learning from museums: Visitor experiences and the making
of meaning. Altamira Press.
Fivush, R., Haden, C. A., & Reese, E. (2006). Elaborating on elaborations: Role of maternal remi-
niscing style in cognitive and socioemotional development. Child Development, 77(6),
1568–1588. https://doi.org/10.1111/j.1467-8624.2006.00960.x
Friedman, A. J. (2007). The extraordinary growth of the science-technology museum. Curator:
The Museum Journal, 50(1), 63–75. https://doi.org/10.1111/j.2151-6952.2007.tb00250.x
Gillies, R. M., & Khan, A. (2009). Promoting reasoned argumentation, problem-solving and
learning during small-group work. Cambridge Journal of Education, 39(1), 7–27. https://doi.
org/10.1080/03057640802701945
Gleason, M. E., & Schauble, L. (1999). Parents’ assistance of their children’s scientific reasoning.
Cognition and Instruction, 17(4), 343–378. https://doi.org/10.1207/S1532690XCI1704_1
Goodwin, M. (2007). Occasioned knowledge exploration in family interaction. Discourse &
Society, 18(1), 93–110. https://doi.org/10.1177/0957926507069459
Gutwill, J. P. (2006). Labels for open-ended exhibits: Using questions and suggestions to motivate
physical activity. Visitor Studies, 9(1), 109.
234 A. M. LAND-ZANDSTRA ET AL.

Gutwill, J. P., & Allen, S. (2009). Facilitating family group inquiry at science museum exhibits.
Science Education, 94(4), 710–742. https://doi.org/10.1002/sce.20387
Haden, C. A. (2010). Talking about science in museums. Child Development Perspectives, 4(1),
62–67. https://doi.org/10.1111/j.1750-8606.2009.00119.x
Haden, C. A., Jant, E. A., Hoffman, P. C., Marcus, M., Geddes, J. R., & Gaskins, S. (2014).
Supporting family conversations and children’s STEM learning in a children’s museum. Early
Childhood Research Quarterly, 29(3), 333–344. https://doi.org/10.1016/j.ecresq.2014.04.004
Hein, G. E. (1998). Learning in the museum. Routledge.
Herrera, Y. M., & Braumoeller, B. F. (2004). Symposium: Discourse and content analysis. Qualitative
Methods, 2(1), 15–19. Retrieved from https://wcfia.harvard.edu/files/wcfia/files/870_symposium.pdf
Hohenstein, J., & Moussouri, T. (2018). Museum learning: Theory and research as tools for
enhancing practice. Routledge.
Hohenstein, J., & Tran, L. U. (2007). Use of questions in exhibit labels to generate explanatory
conversation among science museum visitors. International Journal of Science Education,
29(12), 1557–1580. https://doi.org/10.1080/09500690701494068
Horn, M. S., Phillips, B. C., Evans, E. M., Block, F., Diamond, J., & Shen, C. (2016). Visualizing
biological data in museums: Visitor learning with an interactive tree of life exhibit. Journal of
Research in Science Teaching, 53(6), 895–918. https://doi.org/10.1002/tea.21318
Jant, E. A., Haden, C. A., Uttal, D. H., & Babcock, E. (2014). Conversation and object manipula-
tion influence Children’s Learning in a Museum. Child Development, 85(5), 2029–2045. https://
doi.org/10.1111/cdev.12252
Kisiel, J., Rowe, S., Vartabedian, M. A., & Kopczak, C. (2012). Evidence for family engagement in
scientific reasoning at interactive animal exhibits. Science Education, 96(6), 1047–1070. https://
doi.org/10.1002/sce.21036
Leinhardt, G., Crowley, K., & Knutson, K. (Eds.). (2002). Learning conversations in museums.
Lawrence Erlbaum Associates.
McClain, L. R., & Zimmerman, H. T. (2014). Prior experiences shaping family science conversa-
tions at a nature center. Science Education, 98(6), 1009–1032. https://doi.org/10.1002/sce.21134
National Research Council. (2009). Learning Science in Informal Environments: People, Places, and
Pursuits. Retrieved from http://books.google.com/books?hl=en&lr=&id=vdOIlY7WdJUC&pgis=1
Ornstein, P. A., Haden, C. A., & Hedrick, A. M. (2004). Learning to remember: Social-communi-
cative exchanges and the development of children’s memory skills. Developmental Review,
24(4), 374–395. https://doi.org/10.1016/j.dr.2004.08.004
Packer, J., & Ballantyne, R. (2002). Motivational factors and the visitor experience: A comparison
of three sites. Curator: The Museum Journal, 45(3), 183–198. https://doi.org/10.1111/j.2151-
6952.2002.tb00055.x
Rand, J. (2010). Write and design with family in mind. In D. L. McRainey & J. Russick (Eds.),
Connecting kids to history with museum exhibitions (pp. 257–284). Routledge.
Rogoff, B. (2003). The cultural nature of human development. Oxford University Press.
Rowe, S. (2002). The role of objects in active, distributed meaning-making. In S. G. Paris (Ed.),
Perspectives of object-centered learning in museums2 (pp. 19–35). Lawrence Erlbaum Associates.
Serrell, B. (2015). Exhibit labels: An interpretative approach (Second). Rowman & Littlefield.
Shine, S., & Acosta, T. Y. (2000). Parent-child social play in a children ‘s museum published by :
national council on family relations stable URL : http://www.jstor.org/stable/585700 Linked
references are available on JSTOR for this article : Parent-Child Social Play in a Children’s Mu.
Family Relations, 49(1), 45–52. https://doi.org/10.1111/j.1741-3729.2000.00045.x
Valle, A., & Callanan, M. A. (2006). Similarity comparisons and relational analogies in parent-
child conversations about science topics. Merrill-Palmer Quarterly, 52(1), 96–124. https://doi.
org/10.1353/mpq.2006.0009
Van Gerven, D. J. J., Land-Zandstra, A. M., & Damsma, W. (2018). Authenticity matters: Children
look beyond appearances in their appreciation of museum objects. International Journal of Science
Education, Part B, 8(4), 325–339. https://doi.org/10.1080/21548455.2018.1497218
Vygotsky, L. (1978). Mind in society: The development of higher psychological processes. Harvard
University Press.
 VISITOR STUDIES 235

Zimmerman, C. (2000). The development of scientific reasoning skills. Developmental Review,

20(1), 99–149. https://doi.org/10.1006/drev.1999.0497
Zimmerman, H. T., Perin, S., & Bell, P. (2010). Parents, science and interest: The role of parents
in the development of youths’ interests. Museums & Social Issues, 5(1), 67–86. https://doi.org/
10.1179/msi.2010.5.1.67
Zimmerman, H. T., Reeve, S., & Bell, P. (2008). Distributed expertise in a science center. Journal
of Museum Education, 33(2), 143–152. https://doi.org/10.1080/10598650.2008.11510595

About the authors

Anne M. Land-Zandstra is an Assistant Professor in the Science Communication & Society
Department of Leiden University in the Netherlands. Her research focuses on authenticity in
informal science education, including the role of real objects in science museums.
Kelly Hoefakker was a master’s student in Life Science & Technology and Science
Communication & Society at the time of this research study. She now works as a researcher at
the Netherlands Cancer Institute.
Welmoet Damsma is a Learning Officer and trainer within the Educational Department of
Naturalis Biodiversity Center. As such, she is responsible for research and evaluation and liaises
with different researchers both inside and outside Naturalis. In addition, she develops and organ-
izes professional development of Naturalis’ educators.
236 A. M. LAND-ZANDSTRA ET AL.

Appendix A. Codebook Reasoning with Objects

Category Code Description Typical excerpts

Perceptual talk Identification Pointing at an object, pointing “Look what is lying here”;
out something “What is this?”
Naming Giving an object a name. “Hey, a rock”; “It has a label”;
“a black tablecloth”
Feature Pointing out a specific aspect “Do you see that bump?”; “See
or part those ridges”
Quotation Reading text aloud, referring “What does it look like?”; “This
to label is a real dino egg”
Conceptual talk Simple inference A single interpretative statement “It is an egg”; “This looks like
(reasoning) of interpretation of part of the a rock”
exhibit, based on an
observation.
Complex inference An explicit hypothesis, comparison, “If … , then … .”; “But in those
(reasoning) relationship with knowledge. times there were different
Often using words like because, eggs”; “I think there is a real
therefore, so that, or implying dino inside”
that those words should have
been part of the statement.
Prediction Stating an expectation on what “Maybe a dino will still hedge”
will happen next.
Metacognition Reflection on own state of “Oh, maybe not”; “I thought so,
knowledge; including correcting too”; “I agree”
someone else’s response by
stating the ‘correct’ answer.
Connecting talk Life connection Story, personal association, Stories about chicken at
connection with something grandparents, etc.
familiar (e.g. their own body,
familiar animals)
Knowledge Referring to prior knowledge from “This is an egg”; “Female
connection before their visit to the dinosaurs are larger than
museum; including implicit male ones”; “Dinosaur egg”
reference to prior knowledge by
just stating what they thought
the object was.
Inter-exhibit Link with other elements of the “I think this belongs to Trix
connection exhibition, or the rest of [the T.rex fossil behind
the museum them], otherwise it would
not be here.”; “It could
belong to the one we
saw before”
Strategic talk Use Statement about how to use the “You are not supposed to
exhibit; talk about touching the touch it”; “I wish I brought
object or how to find my magnifying glass to look
the answer at it from up close”
Metaperformance Evaluation of performance “If you drop it, it will break”;
“Now it has our finger
prints”; “We won’t find
the answer”
Affective talk Pleasure Positive statements “Cool”; “Beautiful”
Displeasure Negative statements “Boring”
Intrigue Showing amazement, surprise, “Wow”
fascination