Originality-utility trade-off within creative solutions to the Alternative Uses Task

(1)

Originality-utility trade-off within creative

solutions to the Alternative Uses Task.

Lois van Vliet 10438033

Mentored by Claire Stevenson Bachelor thesis, University of Amsterdam

02-07-2016

Abstract

The aim of this study was to examine the originality-utility trade-off within creative solutions to the alternative Uses Task (AUT) and how originality and utility were related to creativity. Both experts and participants rated the responses from the AUT for originality, utility and creativity. A main

effect for experts ratings has been found between originality and utility; the

originality-utility trade-off. Participants did not show this trade-off. Next, it was found that originality

and utility were good predictors of creativity. Utility and creativity showed a small partial

correlation while originality and creativity showed a high correlation. This indicates that

originality plays a bigger part within the definition of creativity than utility.

(2)

Creativity remains a mysterious phenomenon. How innovative ideas are generated

or how humans come up with new solutions is complex and intriguing (Hennesy and

Amabile, 2010). However, how can creativity be measured? In this study it is attempted to

measure creativity by exploring its two basic characteristics in divergent thinking tasks.

The definition of creativity has been a subject of debate for a while now. According

to Stevenson, Kleibeuker, Drue and Crone (2014), creativity is seen as the ability to generate

ideas, solutions or insights that are novel and executable. While Besemer and O’Quin

(1987) were more concentrated on novelty, resolution (if the product is useful), elaboration

and synthesis (if the product is complex and well-crafted). Stein (1953) stated that creative

work is a novel work that is accepted as useful or satisfying by a group of people in some

point in time. Two main components came back in each interpretation of creativity:

originality and utility. An original idea is novel and personal. It is a vital characteristic of

creativity; when a solution or idea is not novel it will not be considered creative. An useful

or appropriate idea is executable and effective. For example; two creative alternative

solutions for a brick are: a parachute or a doorstop. The first one not useful, while using a

brick as a doorstop is useful. There is no use in thinking of solutions or ideas that can’t be

executed. So an original idea most be useful to be creative. Therefore, utility is next to

originality a vital characteristic. From these characteristics the following definition used and

examined in this paper is: a creative idea must be a novel idea that is also useful and

effective for a group of people.

(3)

originality and utility. This interaction can be explained with the example mentioned above.

Two alternative uses for a brick are; a doorstop or a parachute. A doorstop is useful,

however not original. A brick used as a parachute is original, nonetheless not useful. This

example indicates that there is a trade-off between originality and utility. Whenever an idea

is more original, the less appropriate or useful it becomes and vice versa. This negative

correlation is highly intriguing and can be compared to another known observation:

speed-accuracy trade-off.

The speed-accuracy trade-off is an important notion from reaction time research of

response times on test items (Van der Linden, 2007). This notion indicates that the fraction

of erroneous responses rises towards a lower reaction time. This means that when a subject

has little time, there is a greater possibility to give an inaccurate answer then when the

subject has more time (Schouten and Bekker, 1967; Van der Linden, 2007). Figure 1

illustrate the speed-accuracy trade-off.

(4)

Even though this trade-off is about speed and accuracy, it is a fascinating idea that,

instead of speed and accuracy, there could be a originality-utility trade off. Like the example

with the brick, the more original an idea is, the lower the utility is. Thus, theoretically the

existence of this trade-off is real within the definition of creativity.

The measurement of creativity and the measurement of the correlation between

originality and utility, is another subject of debate within the field of creativity. Basadur,

Runco and Vega (2000) argued that these discussions existed due to different processes

involved in creativity. Two different processes were described: divergent thinking and

convergent thinking. Convergent thinking is the mental process where a subject has to

generate or choose one idea (Gilhooly, 2007; Sivlia, 2015). Divergent thinking is the mental

process where a subject has to generate as many solutions or ideas as they can (Runco and

Acar, 2012; Silvia, 2015; Gilhooly et al, 2007). These different processes ask for different

kind of tasks and measurements. Divergent thinking tasks are considered good measures of

creativity (Silvia, 2015; Runco, 1993; Runco and Acar, 2012). It is a process that widens

thought and that leads to many responses (Silvia, 2015). Stated that divergent thinking is

the marker of creativity (Silvia, 2015), these tasks were used in this study.

There are different kind of divergent thinking tasks where responses are often

evaluated with fluency, originality and flexibility (Runco, 1993; Stevenson et al, 2014;

Ottenmiller et al, 2014; Gilhooly et al, 2007). Many studies do not include a measure of

appropriateness or utility when implementing divergent thinking tasks, despite the fact that

utility is considered a basic characteristic in the standard definition of creativity (Ottenmiller

(5)

characteristics is neglected. Hence, even though creativity and novelty have been examined

before, utility was never a part of these analyses.

To measure originality and utility an alternative uses task (AUT) was done. The AUT is

a typical divergent thinking task, where subjects have to generate as many possible

alternative uses for familiar objects (Ottenmiller et al,2014; Stevenson et al, 2014). The AUT

scores yield various measures for creative responses. Originally this task was made to

measure originality, fluency and feasibility (Mumford, 2001; Ottenmiller et al, 2014;

Stevenson, 2014) not utility.

Since, in this study the correlation between originality an utility

were examined, the responses were rated by their usefulness/appropriateness as well.

Giving previous findings, it was expected that the originality-utility trade-off exists

within the alternative uses task (Stevenson, in press, 2014). Ideas that will score high on

originality will score low on utility. It was also expected that originality and utility are good

predictors for creativity.

Method

Participants

The sample comprised 36 participants (M

age

= 29.12, SD = 12.9, range = 16- 61 years,

74 % females). All participants were recruited through the University of Amsterdam

research system and trough contacted acquaintances. All participants provided informed

consent. All procedures were approved by the Internal Review Board of the University of

Amsterdam Institute of Psychology. After the study, two gift cards were raffled.

(6)

The data was gathered in one week. There were two drop-outs. The number of

subjects used in statistical analyses was 34.

Material

Alternative uses task (AUT). Participants administered a Dutch version of the

computerized AUT (Getzels & Jackson, 1962; Guilford, 1967; Stevenson et al, 2014).

Participants were given the name of an object and they were asked to came up with as

many alternative uses for the object as possible within a 2 min time period (Stevenson et al,

2014). For example, with the target item ‘brick’, a use as ‘doorstop’ might be produced by a

participant. Eight common objects were used in this experiment: brick, fork, paperclip,

towel, book, belt, stick and can. These objects were viewed in this specific order. This order

was equal for each participant.

AUT is traditionally scored for fluency, originality and flexibility. Considering the

standard definition of creativity, originality, utility and creativity were measured. All were

rated on a 10 point scale ( from 1 = ‘not original/useful/creative’ to 10 = ‘very

original/useful/creative’). As well as participants as three experts rated their answers.

The verbal fluency task. With the verbal fluency task, participants were administered

to generate as many animals or jobs as possible within a 2 min time period (Troyer &

Moskovitch & Winocur, 1997). These task was administered as distraction. Hence, these

responses were not taken into account for the analysis

(7)

Procedure

After the participants signed the informed consent they started the experiment. The

experiment consisted of four parts.

1. The participants started with the AUT. The items were presented in a specific order:

brick, fork, paperclip, towel, book, belt, stick and can. After each object the

participants could take a break.

2. After the AUT , the participants had to choose their two most creative solutions for

each item. An example is seen in Figure 2.

Figure 2. An example a selection from one participant for the object book. The

participant had to choose their top two.

3. As a distraction the participants solved two verbal fluency tasks for which they were

asked to think of as many animals or jobs as possible within a 2 minute timeframe.

4. The final task consisted of rating responses on originality, utility and creativity. It was

a 10-points scale, 1 = no originality/utility/creativity and 10 = very

original/useful/creative. The participants were provided with the definitions for

usefulness and originality. For creativity, the participants were asked to use their

own definition. For each item four standard responses were chosen by the experts.

(8)

Each participant had to rate these four responses plus their own top two responses.

An example for these ratings of responses on brick is shown in Figure 3.

Figure 3: example of response ratings for brick. The standard responses and the top two that

were given by a participant were on the left.

Consensual assessment technique

After all responses were gathered, all responses were rated by three experts. The

same scales were used as the ratings of the participants. Hence , 1 = no

originality/utility/creativity and 10 = very original/useful/creative. To reduce bias and

subjectivity there were guidelines for the experts. First, originality was obtained using the

frequency count for each response. A response that was often mentioned by different

participant was rated low, while responses that were unique were rated high. Second, utility

was obtained by rating the plausibility and sensible use. Third, creativity was obtained by

following the standard definition: a creative idea must be a novel idea that is also useful and

effective for a group of people. Finally for each object, all responses must be rated in a

specific order. All responses were first rated by their originality, second utility and last

(9)

creativity. Ratings on the other components were hidden. In such manner bias could been

reduced.

Inter-rater reliability. Inter-rater reliability for the AUT response coding was

established among three experts. R-package irr was used. Agreement and ‘oneway’ was

coded for each response category: creativity, originality and utility. Inter-rater reliability was

low; the ICC was respectively 0.33, 0.47 and 0.227.

Statistical analyses

Correlational study. First, a correlation test between originality and utility ratings,

according to the expert panel, was done. For this correlation test all items and all their

responses were used. Thereafter, the same correlation was calculated for each item

separated.

Multiple linear regression. A multiple linear regression was done for the expert

ratings on creativity, originality and utility. Creativity was the dependent variable. Originality

and utility were the independent variables. The next regression was analyzed:

Creativity = originality + utility + error.

First, this regression was executed for all items and responses. Thereafter, the regression

was executed for all items separately. Together with these regressions, the partial

correlation was calculated. These correlation was calculated with the R-package ppcor.

ANOVA. To compare different participants with different creativity levels to check if

the most creative participants indeed score high on originality and utility, the mean scores

(10)

on creativity for each participant was calculated. Based on these scores the participants

were classified into five groups. These five groups are presented in Table 3. To test whether

these groups differ for creativity, an ANOVA was executed.

Explorative study. Finally the ratings of all participants were examined. This analysis

consisted of calculating correlations between originality and utility. This way, the trade-off

could be studied according to the participants. Following, a regression analysis had been

done to have a better understanding of how participants think about creativity compared to

originality and utility. Again, creativity was the dependent variable and originality and utility

were the independent variables. Similarly to the expert ratings, the correlational study and

the regression analysis were done for each item separately.

Results*

Correlational study. To investigate the originality-utility trade-off, a correlation test

has been done to view how originality and utility were associated.

First, the trade-off overall the ratings was examined. There was a correlation found in the

expected direction, r = - 0.43 , t(1768) = -20.04, p < .05. This correlation is presented in

Figure 4. This figure visualize that when a solution is more original, it is less useful.

(11)

Figure 4. Negative linear correlation between utility and originality, r = 0.43 , t(1768) =

-20.0374, p < .05.

Second, the originality-utility trade-off for each item was calculated separately. The

visual representation of these analysis are seen in Figure 5. The correlation between

originality and utility differed somewhat per item and ranged from -0.10 to -0.70. As can be

seen in Figure 5 ‘paperclip’ had the weakest correlation and ‘belt’ the strongest. All

correlations were significant, as shown in Table 1. All correlations were in the expected

direction. This indicates that within every item an originality-utility trade-off is viewed.

(12)

Figure 5. Visual representation of the correlation between utility and originality for each

item separated.

Multiple regression. A multiple regression was applied to see how well originality and

utility predicted creativity. Creativity was the dependent variable. Utility and originality

were the independent variables. The regression was analyzed in the next form:

(13)

Table 1. Correlations between originality and utility for each item separated.

Originality and utility were good predictors for creativity, R

2

= 0.60, F(2,1768) = 1368, p <

.05. 60 percent of differences in creativity could be explained by originality and utility. The

regression formula associated with these results were:

Creativity = 0.63 * originality + 0.20 * utility + 0.10

A partial correlation was calculated to examine to what extend this regression was

caused by originality and utility. Originality had the strongest partial correlation, r = 0.77 (p <

.05) , compared to utility, r = 0.327 ( p < .05). This means that originality correlated with 0.77

to creativity, considered the effect of utility removed. Utility on the other hand, correlated

with 0.32 to creativity, while the effect of originality was removed. Hence, originality

Correlation

t-value

P-value

Brick

-0.45

t(209) = -7.24

< .05

Can

-0.41

t(204) = -6.43

< .05

Book

-0.44

t(216) = -7.23

< .05

Towel

-0.26

t(282) = -4.48

< .05

Paperclip

-0.19

t(188) = -2.67

< .05

Belt

-0.70

t(199) = -13.76

< .05

Stick

-0.46

t(247) = -8.12

< .05

Fork

-0.39

t(208) = -6.08

< .05

(14)

Figure 6. A visual representation of how creativity, originality and utility are connected.

The previous results are summarized in Figure 6. This 3d plot showed that the data

indicates that there was a relation between originality and utility, originality and creativity,

nonetheless between originality and creativity there was a weak relation; the points were

spread.

The second part of this regression analyses consisted of calculating the originality

slopes and utility slopes for each item separately. These regression analyses and the

corresponding partial correlations are viewed in Table 2. The slopes for originality ranged

from 0.45 to 0.73 with partial correlation between 0.66 and 0.84. The slopes for utility

ranged from 0.08 to 0.33 with corresponding partial correlations between 0.15 to 0.53. It

can be concluded that originality had more influence on the prediction of creativity, than

(15)

Table 2. Regression coefficients and partial correlations for each item separately. The regression line is formed by: creativity = originality + utility + error. The partial correlation consisted of calculated correlations between originality and creativity, and utility and creativity.

Creativity = originality + utility + error Partial correlation Originality Utility

Inter-cept F(df) R

2

p-value Originality and creativity

p –

value Utility and creativity p-value Brick 0.57 0.19 0.47 F(2, 208) = 147.4 0.58 < .05 0.76 < .05 0.38 <.05 Fork 0.56 0.10 0.71 F(2,207) = 187.5 0.64 < .05 0.79 < .05 0.19 < .05 Paperclip 0.45 0.33 0.17 F(2,187) = 91.33 0.49 < .05 0.66 < .05 0.53 < .05 Towel 0.62 0.19 0.05 F(2,281) = 111.7 0.44 < .05 0.66 < .05 0.29 < .05 Book 0.73 0.20 -0.56 F(2,215) = 262.1 0.71 < .05 0.84 < .05 0.37 < .05 Belt 0.68 0.22 -0.35 F(2,198) = 197.5 0.66 < .05 0.78 < .05 0.38 < .05 Stick 0.73 0.08 0.18 F(2,247) = 355.6 0.74 < .05 0.84 < .05 0.15 < .05 Can 0.63 0.27 0.04 F(2,203) = 175 0.63 < .05 0.80 < .05 0.43 < .05

(16)

ANOVA. For a better understanding on how creativity is connected to originality and

utility an ANOVA was executed. All participants were classified in groups of five based on

their mean score for creativity. Considering that these analyses were for exploring creativity

within participants, there were no specific rules for the group classification. The groups

were classified in a way that the size of the groups were close to equal. For each participant

a mean score was calculated for creativity, originality and utility. In Table 3 the group size

and their mean score for creativity was presented.

A one-way ANOVA was executed to check whether the groups significantly differ for

creativity. Creativity was the dependent variable in this analyses. Creativity did significantly

differ between the different groups F(4,29) = 99.62, p< 0 .05. A visual presentation of this

analyses is given in Figure 7. Figure 7 shows the different scores for each participant in a

particular group. As seen, these groups differ for creativity.

Table 3. Group classification, means and group size.

Group

Mean creativity

N

Group 1

<3.5

3 Group 2

> 3.5 and < 4.0

8 Group 3

> 4.0 and < 4.25

10 Group 4

> 4.25 and <4.5

7

(17)

Figure 7. Creativity mean scores of participants in each group.

After dividing participants in different groups, an originality utility plot was made

(Figure 8). In Figure 8 different colors represent the different groups. A trade-off is seen in

this Figure. Hence, the least creative participants score highest on utility and lowest on

originality, while the most creative participants scores high on originality and intermediate

for Utility. It can be stated that the more creative a participant is, the more original their

ideas are, that are sufficient to use.

(18)

Figure 8. Participants in different groups, but with their own means on originality and utility.

Furthermore for each participant a creativity mean score was calculated for each

item. These points are plotted in Figure 9. The participants remained in the same group they

have been previously assigned to. Every color in Figure 9 still represent a different group.

The trade-off was more present in this Figure compared to the previous one. More creative

participants had high scores for originality and intermediate utility scores. It can be stated

that creativity is more connected to originality than utility.

(19)

Figure 9. Participants in different groups, but with their own means on originality and utility

for each item.

Exploratory study. In the exploratory study the participant ratings were used. Each

participant rated four standard responses for creativity, originality and utility. These ratings

were examined.

First the originality-utility trade-off was examined overall the data (Figure 10). Figure

10 shows the correlation of this connection. As can be seen, there is no significant

correlation between the two components ( r = - 0.01, F (1055) = -.37, p = -.71). This indicates

that participants do not think that there is a correlation between these two concepts.

(20)

Figure 10. Originality versus utility overall data, according to participants ( r = - 0.01, F

(1055) = --.37, p = -.71).

Linear regression. After the correlation test a multiple regression analysis was done.

The dependent variable was creativity and the independent variables were originality and

utility. According to the participants, originality and utility were good predictors for

creativity. The next regression was found:

Creativity = 0.71 * originality + 0.18 * utility + 0.32

with R

2

_{= 0.56, F( 2 , 1054) = 683.7, p < .05. It can be implied that 56 percent of differences}

in creativity can be explained by originality and utility. A partial correlation was calculated

to examine to what extend this regression is caused by originality and utility. Originality had

the strongest partial correlation, r = 0.74, p < .05, compared to utility, r = 0.30, p < . 05.

(21)

After using all participants ratings for each item, the originality-utility relation was

examined for each item separately. This trade-off was visualized in Figure 11. As viewed in

this Figure, there seems to be no correlation between originality and utility in different

items. This unexpected result implied that participants don't see an association between

originality and utility.

Furthermore, a regression analysis was done for each item separately. Similar to the

expert ratings; creativity was the dependent variable, originality and utility were the

independent variables. These regression analyses and the corresponding partial correlations

are viewed in Table 4.

(22)

Originality and utility were good predictors to creativity. The regression coefficients

for originality ranged from 0.61 to 0.86 with partial correlation between 0.63 and 0.81. The

regression coefficients for utility ranged from 0.10 to 0.27 with corresponding partial

correlations between 0.02 to 0.41. The only non-significant partial correlation was between

utility and creativity for the item ‘book’.

These results indicated that originality and utility were good predictors for creativity,

nevertheless originality had more influence on the prediction of creativity, than utility.

(23)

Table 4. Regression coefficients and partial correlations for each item separately according to the participants. The regression line is formed by: creativity = originality + utility + error. The partial correlation consisted of calculated correlations between originality and creativity, and utility and creativity.

Creativity = originality + utility + error Partial correlation

Originality Utility

Inter-cept F(df) R

2

p-value Originality and creativity

p –

value Utility and creativity p-value Brick 0.61 0.27 0.46 F(2, 135) = 59.45 0.46 < .05 0.63 _{< .05} 0.41 _<.05 Fork 0.72 0.18 0.41 F(2,129) = 81.98 0.55 < 0.05 0.73 < .05 0.28 < .05 Paperclip 0.63 0.19 0.93 F(2,124) = 60.1 0.48 < .05 0.68 < .05 0.27 < .05 Towel 0.73 0.17 0.29 F(2,129) = 86.2 0.56 < .05 0.75 < .05 0.28 < .05 Book 0.71 0.10 0.83 F(2,127) = 75.78 0.544 < .05 0.72 <.05 0.15 0.08 Belt 0.86 0.22 -0.76 F(2,130) = 129.4 0.66 < .05 0.81 < .05 0.40 < .05 Stick 0.80 0.12 0.24 F(2,131) = 112.5 0.63 < .05 0.78 < .05 0.02 0.021 Can 0.65 0.19 0.28 F(2,128) = 85.19 0.56 < .05 0.73 < .05 0.33 < .05

(24)

Discussion

The aim of the current study was to examine the originality-utility trade-off within

creative solutions to the Alternative Uses task (AUT) and how these variables were related

to creativity. To this end, participants were administered the AUT and the responses were

rated by an expert panel on originality, utility and creativity. Furthermore, participants rated

their own top two creative responses and a number of example responses on originality,

utility and creativity (Figure 3). Results from the expert panel indicated an originality-utility

trade-off. On top of that, results from both participants and the expert panel indicated that

originality an utility were good predictors of creativity.

As expected, the originality-utility trade-off was found for the expert ratings on all

items. To summarize: the more original an alternative solution was, the lower its usefulness.

This trade-off was too presented in each item, nevertheless the slope differed over different

items (Table 2). There were two outliers, ‘belt ‘ and ‘paperclip’. ‘Paperclip’ had a small slope

and ‘belt’ had a high slope. A reason for these findings could be their frequency in

present-day. It is easier to generate non-creative ideas for high frequency objects as compared to

low frequency objects (Forthman & Gerwi & Holling & Celik &strome & Lubart et all, 2014).

‘Paperclip’ was a low frequency object, what led to more idiosyncratic associations and thus

the other hand, was a high frequency object. There were many associations, but many were

shared and banal associations (Forthman et all, 2014). This led to lesser original solutions

that were more useful. Thus, when an item is less frequent in present-day, little solutions

(25)

Another expected result was the goodness of fit of the regression model for

creativity, according to the expert panel. Creativity was predicted by originality and utility.

The creativity definition could be obtained from this regression. Originality had the highest

regression coefficient and partial correlation compared to utility. Thus, it was valid to define

originality as the vital characteristic; a new idea must especially be novel to be creative.

Utility had a small regression coefficient and a small partial correlation. This implied that

utility was not vital. Utility could be seen as an additional feature to creativity. This was

presented in Figure 6, between creativity an utility the points were spread compared to the

points between creativity and originality.

The goodness of fit of the regression model was an expected result for both

participant ratings as for expert ratings. Still there was a difference between the two groups.

Compared to the experts, for the participant the regression coefficient for originality was

higher and the regression coefficient for utility was lower. This indicated that participants

did recognize the importance for originality, nonetheless not for utility in creative thinking.

This could imply that definitions used outside the scientific community differ from

definitions inside the scientific community. For example, during the experiment multiple

participants commented that they did not know the difference between originality and

creativity. This statement was supported by the high regression coefficient (b = 0.71, r = 0.74

) compared to the regression coefficient according to experts ( b = 0.63, r = 0.77). It could be

stated that originality is closer connected to creativity according to the participants than it is

closer to creativity according to the experts.

(26)

This difference in use of definition for creativity and originality could have caused

one unexpected result; the missing originality-utility trade-off within participants. When

originality and creativity are as closely related as the participants think, usefulness could

have been already connected to originality. What led to an unknowingly influence off the

originality and utility ratings. This implies that a novel idea scored higher on originality when

it was also useful. Thus no trade-odd could be found.

Limitations

Some limitations of this study deserve mention and can be informative for future

research. First, the inter-rater reliability was low. This could have been caused by a slight

different scoring pattern. For example: If expert 1 scored consistently 1 point higher than

expert 2, after 1800 responses the ratings differ significantly for just originality, but the

fraction of utility and originality could still be the same. By using means scores, this fraction

is still present and the scores were more objective.

(27)

Another reason for these low inter-rater reliability and for the different results

between expert ratings and participant ratings, can be explained by intrinsic values. To

visualize this problem, Figure 12 shows the different ratings from different participants on

‘Yoga stone’ as an alternative use for ‘Brick’. Every dot represents another participant. It

can be concluded that different participants rated very differently. This could be explained

by different intrinsic values (Kasof, Chem, Himsel & Greenbemer, 2007). Kasof and his

colleagues demonstrated that an individual’s intrinsic values had a big influence on their

creative behavior. For example: participant A thought that the Yoga stone was very original,

nevertheless he did not think it would be comfortable and thus not useful. While participant

B already used a brick as a Yoga stone. Participant B would have rated the yoga stone as not

original but very useful. Participant C on the other hand never thought of using a brick as an

yoga stone, nevertheless thinks it could be very useful.

This example shows that even though the same definitions were used, still

participant A, B and C rated the yoga stone differently. To counter this, mean scores of the

responses could be calculated of the trade- off could be calculated for each person instead

for each item. By analyzing the trade-off for each person you can take the intrinsic values

into account and still examine the trade-off.

.

Another limitation of this study is about the calculated creativity means for each

participant. These means were calculated for every response the participant had. But even

the most creative participant responded with some standard answers: hit, throw, catch. Due

to these responses, creativity means were low. For next researches, these standard answers

(28)

Another manner to counter this is to focus on top two answers. This way, the standard

responses could be filtered out.

The last limitation that should be taken into account for future research is the

randomization of the items. In this study the items were in the same order for every

participant. For more objective results, these items should be random.

(29)

References

Basadur, M. & Runco, M.A.& Vega, L.A. (2000)Understanding How Creative Thinking Skills,

Attitudes and Behaviors Work Together. In Journal of creative behavior. Volume 34, number

2. Besemer, S. P., & O'Quin, K. (1987). Creative product analysis: Testing a model by

developing a judging instrument. In S. G. Isaksen (Ed.), Frontiers of creativity research:

Beyond the basics (pp. 367-389). Buffalo, NY: Bearly.

Cropley, A.J.(2000). Defining and measuring creativity: Are creativity tests worth using?

Roeper Review, 23:2, 72-79.

Forthman, B. & Gerwig, A. & Holling, H. & Celik, P. & Strome, M. & Lubart, T. et all ( 2016).

The be-creative effect in divergent thinking: The interplay of instruction and object

frequency. In: intelligence. Vol.57, pages 25-32

Getzels, J. W., & Jackson, P. W. (1962). Creativity and intelligence: Explorations with gifted

students. New York, NY: Wiley

Gilhooly, K.J. & Fioratou, E.& Anthony, S.H. & Wynn, V. (2007)Divergent thinking: strategies

for generating alternative uses for familiar objects.

British Journal of Psychology, 98, 611–

625. Guilford, J. P. (1967). Creativity: yesterday, today and tomorrow. J. Creat. Behav. 1, 3–14.

doi: 10.1002/j.2162-6057.1967.tb00002.x

(30)

Hennesy, B.A. & Amabile, T.M. (2010) Creativity. In: Annual Review of Psychology. 61,

569-598

Kasof, J.& Chem, C.& Himsel, A.& Greenbemer, E(2007). Values and creativity. Creativity

Research Journal. Volume 19, Issue 2-3.

Linden, van der, W.J. (2007) A hierarchical Framework for modeling speed and Accuracy on

test items. In Psychometrika. Volume 72 No 3, 287-308.

McGraw, K.O., & Wong, S.P. (1996), Forming inferences about some intraclass correlation

coefficients. Psychological Methods, 1, 30-46.

Mumford, M. D. (2001). Something old, something new: Revisiting Guilford’s conception of

creative problem solving. Creativity Research Journal, 13, 267–276.

Nijstad, B.A. & De Dreub, C.K.W. & Rietzschela, E.F. & Baas, M. (2010). The dual

pathway to creativity model: Creative ideation as a function of flexibility and

persistence Bernard A. In European Review of social Psychology. Vol. 21, 34–77

Ottemiller, D.D. & Elliot, C.S. & Giovanneti, T. (2014)Creativity, Overinclusion, and Everyday

Tasks. In Creative research journal. 26(3), 289–296

Schouten, J.F. & Bekker, J.A.M. (1967) Reaction time and accuracy. In Acta Psychologica. 27,

143-153.

Silvia, P.J. (2015) Intelligence and creativity are pretty similar after all. In: Educ Psychol Rev

27:599–606 601

(31)

Stein, M.I. (1953) Creativity and Culture, The Journal of Psychology, 36:2, 311-322.

Stevenson, C.E & Kleibeuker, S.W. & Drue, C.K.W.& Crone, E.A. (2014)Training creative

cognition: adolescence as a flexible period for improving creativity. In Human Neuroscience.

Volume 8 , Article 827.

Runco, M. A. (1993). Divergent thinking, creativity, and giftedness. Gifted Child Quarterly,

37, 16–22.

Runco, M.A.& Acar, S. (2012) Divergent thinking as an indicator for creative potential. In:

Creativity Research Journal, 24:1, 66-75

Runco, C.A. & Jaeger, G.J. (2012). The standard definition of Creativity. In: Creativity

Research. Journal. 24, 92-96.

Runco, M.A. & Acar, A. (2012)Divergent Thinking as an Indicator of Creative Potential.in

Creative Research Journal. 24, 66–75.

Troyer, A.K. & Moskovitch, M. & Winocur, G. (1997) Clustering and Switching as Two

Components of Verbal Fluency: Evidence From Younger and Older Healthy Adults.

Neuropsychology. vol 11. No. 1. 138-146

Zeng, L. & Proctor, R.W.& Salvendy, G. (2011) Can Traditional Divergent Thinking Tests Be

Trusted in Measuring and Predicting Real-World Creativity? In Creative Research Journal. 23,

24–37.

(32)

Appendix A reflectieverslag

A. Omgang met de feedback

In mijn bachelor these project ging de omgang met feedback goed. Wanneer ik de feedback

niet begreep ben ik naar mijn begeleider gestapt zodat ik meer uitleg kon krijgen. De

feedback was daarnaast redelijk. De zaken die mijn begeleider had opgemerkt aan wat er

fout was, was ik het ook bijna altijd mee eens en heb het zo goed mogelijk aangepast.

B. Het project.

Het project ging qua communicatie erg goed. Ik ben goed begeleid door mijn gehele these

en kon altijd bij mijn begeleider terecht. Ook binnen het groepje konden we het goed met

elkaar vinden en voerde iedereen altijd wel de taken uit.

Het enige puntje waar het niet goed ging was de planning. Doordat in het begin het nog een

beetje onduidelijk was wanneer de deadlines waren begonnen we ook erg laat aan het

experimentele gedeelte van het onderzoek. Hierdoor heb ik de laatste maand ontzettend

veel stress ervaren.

Nog een puntje die ik zelf jammer vond, was dat ik uiteindelijk niet heb kunnen helpen bij

het programmeren van het experiment. Ik heb al een klein beetje een achtergrond in

programmeren en had het daarop leerzaam gevonden om het experiment te helpen

programmeren. Helaas door te weinig tijd, is dit niet gelukt.

C. Sterke en zwakke punten.

Sterke punten

Ik denk dat de sterke punten in mijn onderzoek de analyses zijn die ik heb gedaan. Ik heb

elke analyse zelf gecodeerd, waardoor ik zelf het overzicht had van wat er was gedaan en

hoe. Hierdoor heb ik alles goed kunnen interpreteren. Wat ook een sterk punt aan mijn

onderzoek is, is dat een originaliteit-utiliteit onderzoek nooit eerder is uitgevoerd. Dit is het

eerste onderzoek in deze richting, waardoor het een sterke wetenschappelijke relevantie

heeft.

Zwakke punten

De zwakke punten in mijn onderzoek zijn de beoordelingen van de experts en de

deelnemers. Hoewel er zoveel mogelijk is geprobeerd om een bepaalde mate van

(33)

objectiviteit aan te houden, zit er in het beoordelen toch een grote subjectieve factor. Dat is

een minpunt aan dit onderzoek.

Een ander zwakpunt is dat er niet dezelfde uitkomsten waren bij de participanten en de

experts.

Ethische aspecten.

Het experimentele gedeelte was ethisch verantwoord. We waren open naar onze

participanten toe, waardoor er van ethische problemen geen sprake was.

(34)

Appendix B, R-code

# analyse

# inlezen van alle bestanden + sorteren

# alle bestanden van julia baksteen1 = choose.files()

juliabaksteen = read.csv(baksteen1, header=T,sep = ";") order.baksteen<-order(juliabaksteen$autresponse_id) juliabaksteen <- juliabaksteen[order.baksteen,] blik2 = choose.files()

juliablik = read.csv(blik2, header=T,sep = ";") order.juliablik<-order(juliablik$autresponse_id) juliablik <- juliablik[order.juliablik,]

boek3 = choose.files()

juliaboek = read.csv(boek3, header=T,sep = ";") order.boek<-order(juliaboek$autresponse_id) juliaboek <- juliaboek[order.boek,]

juliaboek<-juliaboek[-c(254,255,256),] handdoek4 = choose.files()

juliahanddoek = read.csv(handdoek4, header=T,sep = ";") order.handdoek<-order(juliahanddoek$autresponse_id) juliahanddoek<-juliahanddoek[order.handdoek,] paperclip5 = choose.files()

juliapaperclip = read.csv(paperclip5, header=T,sep = ";") order.paperclip<-order(juliapaperclip$autresponse_id) juliapaperclip<-juliapaperclip[order.paperclip,]

juliapaperclip<-juliapaperclip[-c(229,230),]

riem6 = choose.files()

juliariem = read.csv(riem6, header=T,sep = ";") order.riem<-order(juliariem$autresponse_id) juliariem<-juliariem[order.riem,]

stok7 = choose.files()

juliastok = read.csv(stok7, header=T,sep = ";") order.stok<-order(juliastok$autresponse_id)

(35)

vork8 = choose.files()

juliavork = read.csv(vork8, header=T,sep = ";") order.vork<-order(juliavork$autresponse_id) juliavork<-juliavork[order.vork,]

# alle bestanden van isabella baksteen1 = choose.files()

isabaksteen = read.csv(baksteen1, header=T,sep = ";") order.baksteen<-order(isabaksteen$autresponse_id) isabaksteen <- isabaksteen[order.baksteen,]

blik2 = choose.files()

isablik = read.csv(blik2, header=T,sep = ";") order.isablik<-order(isablik$autresponse_id) isablik <- isablik[order.isablik,]

isaboek = read.csv(boek3, header=T,sep = ";") order.boek<-order(isaboek$autresponse_id) isaboek <- isaboek[order.boek,]

handdoek4 = choose.files()

isahanddoek = read.csv(handdoek4, header=T,sep = ";") order.handdoek<-order(isahanddoek$autresponse_id) isahanddoek<-isahanddoek[order.handdoek,]

paperclip5 = choose.files()

isapaperclip = read.csv(paperclip5, header=T,sep = ";") order.paperclip<-order(isapaperclip$autresponse_id) isapaperclip<-isapaperclip[order.paperclip,]

isariem = read.csv(riem6, header=T,sep = ";") order.riem<-order(isariem$autresponse_id) isariem<-isariem[order.riem,]

isastok = read.csv(stok7, header=T,sep = ";") order.stok<-order(isastok$autresponse_id) isastok<-isastok[order.stok,]

isavork = read.csv(vork8, header=T,sep = ";") order.vork<-order(isavork$autresponse_id) isavork<-isavork[order.vork,]

(36)

# ale bestanden van lois baksteen1 = choose.files()

loisbaksteen = read.csv(baksteen1, header=T,sep = ";") order.baksteen<-order(loisbaksteen$autresponse_id) loisbaksteen <- loisbaksteen[order.baksteen,]

blik2 = choose.files()

loisblik = read.csv(blik2, header=T,sep = ";") order.loisblik<-order(loisblik$autresponse_id) loisblik <- loisblik[order.loisblik,]

loisboek = read.csv(boek3, header=T,sep = ";") order.boek<-order(loisboek$autresponse_id) loisboek <- loisboek[order.boek,]

handdoek4 = choose.files()

loishanddoek = read.csv(handdoek4, header=T,sep = ";") order.handdoek<-order(loishanddoek$autresponse_id) loishanddoek<-loishanddoek[order.handdoek,]

loishanddoek<-loishanddoek[-307,] paperclip5 = choose.files()

loispaperclip = read.csv(paperclip5, header=T,sep = ";") order.paperclip<-order(loispaperclip$autresponse_id) loispaperclip<-loispaperclip[order.paperclip,]

loisriem = read.csv(riem6, header=T,sep = ";") order.riem<-order(loisriem$autresponse_id) loisriem<-loisriem[order.riem,]

loisstok = read.csv(stok7, header=T,sep = ";") order.stok<-order(loisstok$autresponse_id) loisstok<-loisstok[order.stok,]

loisvork = read.csv(vork8, header=T,sep = ";") order.vork<-order(loisvork$autresponse_id) loisvork<-loisvork[order.vork,]

# Per item niet goede antwoorden eruit halen, dan krijg je vervolgens de totale data. # baksteen

(37)

validbaksteen invalidbaksteen=c(NA) for (i in 1:length(validbaksteen)){ if (validbaksteen[i]!= 0){ invalidbaksteen = c(invalidbaksteen, i) }} invalidbaksteen

baksteen<-data.frame(isabaksteen[,1:8], loisbaksteen[,5:7], juliabaksteen[,5:7]) baksteen<-baksteen[-c(invalidbaksteen[2:length(invalidbaksteen)]),] # blikje validblik = isablik[,8]+loisblik[,8]+juliablik[,8] invalidblik=c(NA) for (i in 1:length(validblik)){ if (validblik[i]!= 0){ invalidblik = c(invalidblik, i) }} invalidblik

blik<-data.frame(isablik[,1:8], loisblik[,5:7], juliablik[,5:7]) blik<-blik[-c(invalidblik[2:length(invalidblik)]),] # boek validboek = isaboek[,8]+loisboek[,8]+juliaboek[,8] invalidboek=c(NA) for (i in 1:length(validboek)){ if (validboek[i]!= 0){ invalidboek = c(invalidboek, i) }} invalidboek

boek<-data.frame(isaboek[,1:8], loisboek[,5:7], juliaboek[,5:7]) boek<-boek[-c(invalidboek[2:length(invalidboek)]),]

(38)

validhanddoek = isahanddoek[,8]+loishanddoek[,8]+juliahanddoek[,8] invalidhanddoek=c(NA) for (i in 1:length(validhanddoek)){ if (validhanddoek[i]!= 0){ invalidhanddoek = c(invalidhanddoek, i) }} invalidhanddoek

handdoek<-data.frame(isahanddoek[,1:8], loishanddoek[,5:7], juliahanddoek[,5:7]) handdoek<-handdoek[-c(invalidhanddoek[2:length(invalidhanddoek)]),] # paperclip validpaperclip = isapaperclip[,8]+loispaperclip[,8]+juliapaperclip[,8] invalidpaperclip=c(NA) for (i in 1:length(validpaperclip)){ if (validpaperclip[i]!= 0){ invalidpaperclip = c(invalidpaperclip, i) }} invalidpaperclip

paperclip<-data.frame(isapaperclip[,1:8], loispaperclip[,5:7], juliapaperclip[,5:7]) paperclip<-paperclip[-c(invalidpaperclip[2:length(invalidpaperclip)]),] # riem validriem = isariem[,8]+loisriem[,8]+juliariem[,8] invalidriem=c(NA) for (i in 1:length(validriem)){ if (validriem[i]!= 0){ invalidriem = c(invalidriem, i) }} invalidriem

riem<-data.frame(isariem[,1:8], loisriem[,5:7], juliariem[,5:7]) riem<-riem[-c(invalidriem[2:length(invalidriem)]),]

# stok

(39)

invalidstok=c(NA) for (i in 1:length(validstok)){ if (validstok[i]!= 0){ invalidstok = c(invalidstok, i) }} invalidstok

stok<-data.frame(isastok[,1:8], loisstok[,5:7], juliastok[,5:7]) stok<-stok[-c(invalidstok[2:length(invalidstok)]),] names(employ.data)[3] <- 'firstday' names(stok)[11]<- 'creativity.1' names(stok)[14]<- 'creativity.2' # vork validvork = isavork[,8]+loisvork[,8]+juliavork[,8] invalidvork=c(NA) for (i in 1:length(validvork)){ if (validvork[i]!= 0){ invalidvork = c(invalidvork, i) }} invalidvork

vork<-data.frame(isavork[,1:8], loisvork[,5:7], juliavork[,5:7]) vork<-vork[-c(invalidvork[2:length(invalidvork)]),] names(stok) # --- mydata<-rbind(baksteen,blik,boek,handdoek,paperclip,riem,stok,vork) #--- # de analyses

# is er een verschil tussen de experts? # originaliteit

#ANOVA toets

(40)

originaliteit = data.frame(mydata$originality, mydata$originality.1, mydata$originality.2) utility = data.frame(mydata$utility, mydata$utility.1, mydata$utility.2)

creativiteit = data.frame(mydata$creativity, mydata$creativity.1, mydata$creativity.2) # de gemiddeldes berekenen per variabele.

originaliteit = apply(originaliteit,1,mean) utility = apply(utility,1,mean)

creativiteit= apply(creativiteit,1,mean)

#--- # grafieken tegen elkaar uitzetten

old.par <- par(mfrow=c(1,1))

plot(originaliteit,utility , main = "Originality and utility overall", type = "p", xlab = "Originality", ylab = "Utility", col = 1, bg = 'white',pch = 19 )

abline(lm(originaliteit~utility)) cor.test(originaliteit, utility) fit fit<-lm(originaliteit~utility) summary(fit) # 3D Scatterplot library(rgl)

plot3d(utility,originaliteit,creativiteit,pch=19, cex = 2.5,NAcol = "grey", highlight.3d = TRUE, xlab = 'Utility', ylab= 'Originality', zlab = 'Creativity', cex.axis = 2.5)

fit<-lm(creativiteit~originaliteit+utility) summary(fit)

fit ?plot3d

# utility vs creativity

plot(utility,creativiteit,main = "creativity and utility", pch=19,type = "p", xlab = "Utility", ylab = "Creativiteit", col = "black")

abline(lm(creativiteit~utility)) fit2<-lm(creativiteit~utility) summary(fit2)

(41)

plot(originaliteit,creativiteit,las=1,main = "Originality and Creativity", pch = 19, type = "p", xlab = "Originality", ylab = "Creativity", col = "black")

abline(lm(creativiteit~originaliteit)) fit<-lm(creativiteit~originaliteit) summary(fit) fit #regressieanalyse #--- # originaliteit tegenover utility voor elk item #--- # baksteen

old.par <- par(mfrow=c(3, 3)) #baksteen

baksteen.originaliteit = data.frame(baksteen$originality, baksteen$originality.1, baksteen$originality.2)

baksteen.utility = data.frame(baksteen$utility, baksteen$utility.1, baksteen$utility.2)

baksteen.creativiteit = data.frame(baksteen$creativity, baksteen$creativity.1, baksteen$creativity.2) # de gemiddeldes berekenen per variabele.

originaliteitbaksteen = apply(baksteen.originaliteit,1,mean) utilitybaksteen = apply(baksteen.utility,1,mean)

creativiteitbaksteen= apply(baksteen.creativiteit,1,mean) # grafieken tegen elkaar uitzetten

plot(originaliteitbaksteen,utilitybaksteen, pch = 19, main = "Brick", type = "p", xlab = "Originality", ylab = "Utility", col = "black")

cor(originaliteitbaksteen,utilitybaksteen) (cor.test(originaliteitbaksteen, utilitybaksteen))

abline(lm(originaliteitbaksteen~utilitybaksteen),col = 2) lm(originaliteitbaksteen~utilitybaksteen)

(42)

fit3

summary(fit3)

# blikje

blik.originaliteit = data.frame(blik$originality, blik$originality.1, blik$originality.2) blik.utility = data.frame(blik$utility, blik$utility.1, blik$utility.2)

blik.creativiteit = data.frame(blik$creativity, blik$creativity.1, blik$creativity.2) # de gemiddeldes berekenen per variabele.

originaliteitblik = apply(blik.originaliteit,1,mean) utilityblik = apply(blik.utility,1,mean)

creativiteitblik= apply(blik.creativiteit,1,mean) # grafieken tegen elkaar uitzetten

plot(originaliteitblik,utilityblik, pch = 19, main = "Can", type = "p", xlab = "Originality", ylab = "Utility", col = "black") cor(originaliteitblik,utilityblik) cor.test(originaliteitblik,utilityblik) abline(lm(originaliteitblik~utilityblik),col = 2) lm(originaliteitblik~utilityblik) summary(lm(originaliteitblik~utilityblik))

fit3<-lm(creativiteitblik ~ originaliteitblik + utilityblik) fit3

summary(fit3) # boek

boek.originaliteit = data.frame(boek$originality, boek$originality.1, boek$originality.2) boek.utility = data.frame(boek$utility, boek$utility.1, boek$utility.2)

boek.creativiteit = data.frame(boek$creativity, boek$creativity.1, boek$creativity.2) # de gemiddeldes berekenen per variabele.

originaliteitboek = apply(boek.originaliteit,1,mean) utilityboek = apply(boek.utility,1,mean)

creativiteitboek= apply(boek.creativiteit,1,mean) # grafieken tegen elkaar uitzetten

plot(originaliteitboek,utilityboek, pch = 19, main = "Book", type = "p", xlab = "Originality", ylab = "Utility", col = "black")

cor(originaliteitboek,utilityboek) cor.test(originaliteitboek,utilityboek)

(43)

lm(originaliteitboek~utilityboek)

summary(lm(originaliteitboek~utilityboek))

fit3<-lm(creativiteitboek ~ originaliteitboek + utilityboek) fit3

summary(fit3)

# handdoek

handdoek.originaliteit = data.frame(handdoek$originality, handdoek$originality.1, handdoek$originality.2)

handdoek.utility = data.frame(handdoek$utility, handdoek$utility.1, handdoek$utility.2) handdoek.creativiteit = data.frame(handdoek$creativity, handdoek$creativity.1,

handdoek$creativity.2)

# de gemiddeldes berekenen per variabele.

originaliteithanddoek = apply(handdoek.originaliteit,1,mean) utilityhanddoek = apply(handdoek.utility,1,mean)

creativiteithanddoek= apply(handdoek.creativiteit,1,mean) # grafieken tegen elkaar uitzetten

plot(originaliteithanddoek,utilityhanddoek, pch = 19, main = "Towel", type = "p", xlab = "Originality", ylab = "Utility", col = "black")

cor(originaliteithanddoek,utilityhanddoek) cor.test(originaliteithanddoek,utilityhanddoek)

abline(lm(originaliteithanddoek~utilityhanddoek),col = 2) lm(originaliteithanddoek~utilityhanddoek)

summary(lm(originaliteithanddoek~utilityhanddoek))

fit3<-lm(creativiteithanddoek ~ originaliteithanddoek + utilityhanddoek) fit3

summary(fit3) # paperclip

paperclip.originaliteit = data.frame(paperclip$originality, paperclip$originality.1, paperclip$originality.2)

paperclip.utility = data.frame(paperclip$utility, paperclip$utility.1, paperclip$utility.2) paperclip.creativiteit = data.frame(paperclip$creativity, paperclip$creativity.1, paperclip$creativity.2)

# de gemiddeldes berekenen per variabele.

originaliteitpaperclip = apply(paperclip.originaliteit,1,mean) utilitypaperclip = apply(paperclip.utility,1,mean)

(44)

# grafieken tegen elkaar uitzetten

plot(originaliteitpaperclip,utilitypaperclip,pch = 19, main = "Paperclip", type = "p", xlab = "Originality", ylab = "Utility", col = "black")

cor.test(originaliteitpaperclip,utilitypaperclip)

abline(lm(originaliteitpaperclip~utilitypaperclip),col = 2) lm(originaliteitpaperclip~utilitypaperclip)

summary(lm(originaliteitpaperclip~utilitypaperclip))

fit3<-lm(creativiteitpaperclip ~ originaliteitpaperclip + utilitypaperclip) fit3

summary(fit3)

# riem

riem.originaliteit = data.frame(riem$originality, riem$originality.1, riem$originality.2) riem.utility = data.frame(riem$utility, riem$utility.1, riem$utility.2)

riem.creativiteit = data.frame(riem$creativity, riem$creativity.1, riem$creativity.2) # de gemiddeldes berekenen per variabele.

originaliteitriem = apply(riem.originaliteit,1,mean) utilityriem = apply(riem.utility,1,mean)

creativiteitriem= apply(riem.creativiteit,1,mean) # grafieken tegen elkaar uitzetten

plot(originaliteitriem,utilityriem, pch = 19, main = "Belt", type = "p", xlab = "Originality", ylab = "Utility", col = "black")

cor.test(originaliteitriem,utilityriem)

abline(lm(originaliteitriem~utilityriem),col = 2) lm(originaliteitriem~utilityriem)

summary(lm(originaliteitriem~utilityriem))

fit3<-lm(creativiteitriem ~ originaliteitriem + utilityriem) fit3

summary(fit3)

# stok

stok.originaliteit = data.frame(stok$originality, stok$originality.1, stok$originality.2) stok.utility = data.frame(stok$utility, stok$utility.1, stok$utility.2)

stok.creativiteit = data.frame(stok$creativity, stok$creativity.1, stok$creativity.2) # de gemiddeldes berekenen per variabele.

originaliteitstok = apply(stok.originaliteit,1,mean) utilitystok = apply(stok.utility,1,mean)

(45)

# grafieken tegen elkaar uitzetten

plot(originaliteitstok,utilitystok,pch = 19, main = "Stick", type = "p", xlab = "Originality", ylab = "Utility", col = "black")

cor.test(originaliteitstok,utilitystok)

abline(lm(originaliteitstok~utilitystok),col = 2) lm(originaliteitstok~utilitystok)

summary(lm(originaliteitstok~utilitystok))

fit3<-lm(creativiteitstok ~ originaliteitstok + utilitystok) fit3

summary(fit3)

# vork

vork.originaliteit = data.frame(vork$originality, vork$originality.1, vork$originality.2) vork.utility = data.frame(vork$utility, vork$utility.1, vork$utility.2)

vork.creativiteit = data.frame(vork$creativity, vork$creativity.1, vork$creativity.2) # de gemiddeldes berekenen per variabele.

originaliteitvork = apply(vork.originaliteit,1,mean) utilityvork = apply(vork.utility,1,mean)

creativiteitvork= apply(vork.creativiteit,1,mean) # grafieken tegen elkaar uitzetten

plot(originaliteitvork,utilityvork, pch =19, main = "Fork", type = "p", xlab = "Originality", ylab = "Utility", col = "black")

cor.test(originaliteitvork,utilityvork)

abline(lm(originaliteitvork~utilityvork),col = 2) lm(originaliteitvork~utilityvork)

summary(lm(originaliteitvork~utilityvork))

fit3<-lm(creativiteitvork ~ originaliteitvork + utilityvork) fit3

summary(fit3)

(46)

# baksteen

old.par <- par(mfrow=c(1,1))

#analyse voor het maken van proefpersonen. Hierdoor zouden we kunnen kijken hoe verschillende personen dus hebben gescoord.

# eerste nemen we baksteen

baksteen.nu<-data.frame(baksteen$autresponse_id,baksteen$participant_fk, baksteen$ response,creativiteitbaksteen)

order.ppn.baksteen<-order(baksteen.nu$baksteen.participant_fk) baksteen.nu<-baksteen.nu[order.ppn.baksteen,]

# moet elk proefpersoon een gemiddelde creativiteit krijgen. # hier komt per proefpersoon het gemiddelde.

ppn.gemiddelde.baksteen <- data.frame( nummer = numeric(36), proefpersoon= numeric(36), Gemiddelde.baksteen = numeric(36), stringsAsFactors=FALSE ) j=1 som=0 l=0 ppn=1 for (i in 1:nrow(baksteen.nu)){ if (baksteen.nu[i,2]==ppn){

som = som + baksteen.nu$creativiteitbaksteen[i] l=l+1 ppn.gemiddelde.baksteen[j,3]=som/l ppn.gemiddelde.baksteen[j,2]=ppn ppn.gemiddelde.baksteen[j,1]=j} else{ ppn.gemiddelde.baksteen[j,3]=som/l ppn.gemiddelde.baksteen[j,2]=ppn ppn.gemiddelde.baksteen[j,1]=j ppn=baksteen.nu[i,2] j=j+1 som = baksteen.nu$creativiteitbaksteen[i] l=1 } } ppn.gemiddelde.baksteen<-ppn.gemiddelde.baksteen[-c(25,36),]

(47)

# blikje # blik

# eerste nemen we blik

blik.nu<-data.frame(blik$autresponse_id,blik$participant_fk, blik$ response,creativiteitblik) order.ppn.blik<-order(blik.nu$blik.participant_fk)

blik.nu<-blik.nu[order.ppn.blik,]

ppn.gemiddelde.blik <- data.frame( nummer = numeric(36), proefpersoon= numeric(36), Gemiddelde.blik = numeric(36), stringsAsFactors=FALSE ) j=1 som=0 l=0 ppn=1 for (i in 1:nrow(blik.nu)){ if (blik.nu[i,2]==ppn){

som = som + blik.nu$creativiteitblik[i] l=l+1 ppn.gemiddelde.blik[j,3]=som/l ppn.gemiddelde.blik[j,2]=ppn ppn.gemiddelde.blik[j,1]=j} else{ ppn.gemiddelde.blik[j,3]=som/l ppn.gemiddelde.blik[j,2]=ppn ppn.gemiddelde.blik[j,1]=j ppn=blik.nu[i,2] j=j+1 som = blik.nu$creativiteitblik[i] l=1 } }

(48)

# boek

# eerste nemen we boek

boek.nu<-data.frame(boek$autresponse_id,boek$participant_fk, boek$ response,creativiteitboek) order.ppn.boek<-order(boek.nu$boek.participant_fk)

boek.nu<-boek.nu[order.ppn.boek,]

ppn.gemiddelde.boek <- data.frame( nummer = numeric(36), proefpersoon= numeric(36), Gemiddelde.boek = numeric(36), stringsAsFactors=FALSE ) j=1 som=0 l=0 ppn=1 for (i in 1:nrow(boek.nu)){ if (boek.nu[i,2]==ppn){

som = som + boek.nu$creativiteitboek[i] l=l+1 ppn.gemiddelde.boek[j,3]=som/l ppn.gemiddelde.boek[j,2]=ppn ppn.gemiddelde.boek[j,1]=j} else{ ppn.gemiddelde.boek[j,3]=som/l ppn.gemiddelde.boek[j,2]=ppn ppn.gemiddelde.boek[j,1]=j ppn=boek.nu[i,2] j=j+1 som = boek.nu$creativiteitboek[i] l=1 } }

insertRow <- function(existingDF, newrow, r) {

(49)

existingDF }

nrow(ppn.gemiddelde.boek)

ppn.gemiddelde.boek<-insertRow(ppn.gemiddelde.boek,c(11,22,0),11) ppn.gemiddelde.boek<-ppn.gemiddelde.boek[-c(35,36,37),]

# eerste nemen we handdoek

handdoek.nu<-data.frame(handdoek$autresponse_id,handdoek$participant_fk, handdoek$ response,creativiteithanddoek)

order.ppn.handdoek<-order(handdoek.nu$handdoek.participant_fk) handdoek.nu<-handdoek.nu[order.ppn.handdoek,]

ppn.gemiddelde.handdoek <- data.frame( nummer = numeric(36), proefpersoon= numeric(36), Gemiddelde.handdoek = numeric(36), stringsAsFactors=FALSE ) j=1 som=0 l=0 ppn=1 for (i in 1:nrow(handdoek.nu)){ if (handdoek.nu[i,2]==ppn){

som = som + handdoek.nu$creativiteithanddoek[i] l=l+1 ppn.gemiddelde.handdoek[j,3]=som/l ppn.gemiddelde.handdoek[j,2]=ppn ppn.gemiddelde.handdoek[j,1]=j} else{ ppn.gemiddelde.handdoek[j,3]=som/l ppn.gemiddelde.handdoek[j,2]=ppn ppn.gemiddelde.handdoek[j,1]=j ppn=handdoek.nu[i,2] j=j+1 som = handdoek.nu$creativiteithanddoek[i] l=1 }

(50)

ppn.gemiddelde.handdoek<-ppn.gemiddelde.handdoek[-c(19,36),] # paperclip paperclip.nu<-data.frame(paperclip$autresponse_id,paperclip$participant_fk, paperclip$ response,creativiteitpaperclip) order.ppn.paperclip<-order(paperclip.nu$paperclip.participant_fk) paperclip.nu<-paperclip.nu[order.ppn.paperclip,]

ppn.gemiddelde.paperclip <- data.frame( nummer = numeric(36), proefpersoon= numeric(36), Gemiddelde.paperclip = numeric(36), stringsAsFactors=FALSE ) j=1 som=0 l=0 ppn=1 for (i in 1:nrow(paperclip.nu)){ if (paperclip.nu[i,2]==ppn){

som = som + paperclip.nu$creativiteitpaperclip[i] l=l+1 ppn.gemiddelde.paperclip[j,3]=som/l ppn.gemiddelde.paperclip[j,2]=ppn ppn.gemiddelde.paperclip[j,1]=j} else{ ppn.gemiddelde.paperclip[j,3]=som/l ppn.gemiddelde.paperclip[j,2]=ppn ppn.gemiddelde.paperclip[j,1]=j ppn=paperclip.nu[i,2] j=j+1 som = paperclip.nu$creativiteitpaperclip[i] l=1 } } ppn.gemiddelde.paperclip<-ppn.gemiddelde.paperclip[-c(35,36),] # riem

(51)

order.ppn.riem<-order(riem.nu$riem.participant_fk) riem.nu<-riem.nu[order.ppn.riem,]

ppn.gemiddelde.riem <- data.frame( nummer = numeric(36), proefpersoon= numeric(36), Gemiddelde.riem = numeric(36), stringsAsFactors=FALSE ) j=1 som=0 l=0 ppn=1 for (i in 1:nrow(riem.nu)){ if (riem.nu[i,2]==ppn){

som = som + riem.nu$creativiteitriem[i] l=l+1 ppn.gemiddelde.riem[j,3]=som/l ppn.gemiddelde.riem[j,2]=ppn ppn.gemiddelde.riem[j,1]=j} else{ ppn.gemiddelde.riem[j,3]=som/l ppn.gemiddelde.riem[j,2]=ppn ppn.gemiddelde.riem[j,1]=j ppn=riem.nu[i,2] j=j+1 som = riem.nu$creativiteitriem[i] l=1 } } ppn.gemiddelde.riem<-ppn.gemiddelde.riem[-c(35,36),] # eerste nemen we stok

stok.nu<-data.frame(stok$autresponse_id,stok$participant_fk, stok$ response,creativiteitstok) order.ppn.stok<-order(stok.nu$stok.participant_fk)

stok.nu<-stok.nu[order.ppn.stok,]

(52)

nummer = numeric(36), proefpersoon= numeric(36), Gemiddelde.stok = numeric(36), stringsAsFactors=FALSE ) j=1 som=0 l=0 ppn=1 for (i in 1:nrow(stok.nu)){ if (stok.nu[i,2]==ppn){

som = som + stok.nu$creativiteitstok[i] l=l+1 ppn.gemiddelde.stok[j,3]=som/l ppn.gemiddelde.stok[j,2]=ppn ppn.gemiddelde.stok[j,1]=j} else{ ppn.gemiddelde.stok[j,3]=som/l ppn.gemiddelde.stok[j,2]=ppn ppn.gemiddelde.stok[j,1]=j ppn=stok.nu[i,2] j=j+1 som = stok.nu$creativiteitstok[i] l=1 } } ppn.gemiddelde.stok<-ppn.gemiddelde.stok[-c(35,36),] # vork

vork.nu<-data.frame(vork$autresponse_id,vork$participant_fk, vork$ response,creativiteitvork) order.ppn.vork<-order(vork.nu$vork.participant_fk)

vork.nu<-vork.nu[order.ppn.vork,]

ppn.gemiddelde.vork <- data.frame( nummer = numeric(36),

proefpersoon= numeric(36), Gemiddelde.vork = numeric(36), stringsAsFactors=FALSE

(53)

j=1 som=0 l=0 ppn=1 for (i in 1:nrow(vork.nu)){ if (vork.nu[i,2]==ppn){

som = som + vork.nu$creativiteitvork[i] l=l+1 ppn.gemiddelde.vork[j,3]=som/l ppn.gemiddelde.vork[j,2]=ppn ppn.gemiddelde.vork[j,1]=j} else{ ppn.gemiddelde.vork[j,3]=som/l ppn.gemiddelde.vork[j,2]=ppn ppn.gemiddelde.vork[j,1]=j ppn=vork.nu[i,2] j=j+1 som = vork.nu$creativiteitvork[i] l=1 } } ppn.gemiddelde.vork<-ppn.gemiddelde.vork[-c(25,36),] ppn.gemiddelde.alles<-data.frame(ppn.gemiddelde.baksteen,ppn.gemiddelde.blik[,3],ppn.gemiddelde.boek[,3],ppn.gemidd elde.handdoek[,3],ppn.gemiddelde.paperclip[,3],ppn.gemiddelde.riem[,3],ppn.gemiddelde.stok[,3], ppn.gemiddelde.vork[,3]) mean<-apply(ppn.gemiddelde.alles[,3:10],1,mean) ppn.gemiddelde.alles<-data.frame(ppn.gemiddelde.alles,mean)

# groepen maken van de proefpersonen hist(mean) group=numeric(34) group[(mean>0)&(mean<3.6)]=1 group[(mean>3.59)&(mean<4.0)]=2 group[(mean>3.99)&(mean<4.25)]=3 group[(mean>4.24)&(mean<4.5)]=4 group[(mean>4.48)]= 5

(54)

hist(group) group1<-subset(ppn.gemiddelde.alles,group==1) group2<-subset(ppn.gemiddelde.alles,group==2) group3<-subset(ppn.gemiddelde.alles,group==3) group4<-subset(ppn.gemiddelde.alles,group==4) group5<-subset(ppn.gemiddelde.alles,group==5)

# deze groepen vergelijken met elkaar

# --- # alles voor elke proefpersoon verschillend maken. # baksteen

# eerste nemen we baksteen

baksteen.originaliteit<-data.frame(baksteen$autresponse_id,baksteen$participant_fk, baksteen$ response,originaliteitbaksteen)

order.ppn.baksteen<-order(baksteen.originaliteit$baksteen.participant_fk) baksteen.originaliteit<-baksteen.originaliteit[order.ppn.baksteen,]

# moet elk proefpersoon een gemiddelde originaliteit krijgen. # hier komt per proefpersoon het gemiddelde.

ppn.originaliteit.baksteen <- data.frame( nummer = numeric(36),

proefpersoon= numeric(36),

Gemiddelde.baksteen = numeric(36), stringsAsFactors=FALSE