Introduction

According to Simpson and Weiner (1989, p. 486), humor is “the faculty of perceiving what is ludicrous or amusing or of expressing it in speech, writing, or other composition; jocose imagination or treatment of a subject.” According to psychological point of view, humor has four components. The first one is ‘the social context of humor,’ which refers to this point that humor could occur in any social setting such as strangers and friends. The other one refers to ‘cognitive-perceptual processes in humor’. This component refers to this point that human-in order to perceive and produce humor-should process information received from the environment and memory, mentally. The third component is ‘emotional aspects of humor.’ The final case refers to the vocal-behavioral expression of laughter. Their social purpose is to motivate others to behave in particular ways.

     On the other hand, according to previous studies, learners to experience successful language acquisition require extensive word knowledge (Heidari, 2019). They are unable to increase their word knowledge including collocations due to insufficient meaningful input in the language classroom. Wagner and Urios-Aparisi (2011) believe that in language teaching/learning, using humor in classroom promotes immediacy, satisfaction, agreement, and motivation. They believe that humor can be a tool for encouraging learners in language acquisition as well as cultural knowledge. Generally speaking, it is clear that the use of humor in the classroom helps learners to reduce tension, stress, anxiety, and boredom; improves the relationship between students and teachers, makes the classroom less threatening and more enjoyable and creates positive attitudes for learning. Collocation is known as an important element for EFL learners in learning vocabulary. Learners in order to improve their communication, comprehension, and other skills in a foreign language need to learn collocations. Therefore, collocation should be added to teaching curriculums in order to help EFL learners improve a foreign language. One way for helping learners improve a foreign language is the use of humor texts which is known as an element for reducing some factors which have negative effects on EFL learners. So, in this study, I will consider the effect of reading humor texts on learning collocations by male and female intermediate learners.

     Because of the importance of collocation for improvement of EFL learners’ knowledge of collocation, one way for helping learners to improve a foreign language may be is the use of humor texts in the curriculums. The study, in fact, addresses this research question:

What is the effect of reading humor texts on learning collocation by    Iranian EFL learners?

Literature Review

Humor

Most of researchers believe that humor involves an idea, image, text, or event that is in some sense incongruous, odd, unusual, unexpected, surprising, or out of the ordinary. Based on Rashidi, et al. (2014) study, humor text motivates learners in the classroom. So, creative teachers to motivate their students to a matter of a language can use humor text. It is clear that the reaction of students different from each other because of their various cultures, familiarity with the format of humor text, personality, intelligence, etc.

     Huss and Eastep (2016) studied the attitudes of university faculty toward humor as a pedagogical tool. In this study, teachers answer to a questionnaire which consider about their opinion towards the use of humor as a pedagogical tool. According to the result of this study, instructors think that humor has positive effects in their planning in traditional and web-based classes. So they say that humor is a part of their plans. One positive effect of the use of humor is to create an atmosphere to students to be relaxed, and create an enjoyable classroom for them.

     The article by Hayati, Gooniband, and Shakeri (2011, pp.652-661) studied “using humorous texts in improving reading comprehension of EFL learners.” They talk about features of humor that it is a sole, universal part of humans’ experience and is expressed by language. The use of humor in the context of second language has great advantages in teaching/learning a language. They refer to this point that learners have not motivation and interest for reading a text, so they have weak memory storage of the reading passage. They collect forty students randomly and divide them into two groups: experimental and control group. The experimental group receives humorous text but the other group studied the same reading text without a humorous concept. Finally, they know that humorous group improved in the recall and comprehension of the text.

Vaezi and Fallah (2012) points to the importance of humor in human’s life and its effect on ESL students’ learning. The aim of this study is to find the opinions and reflections of graduate students for the use of humor as pedagogical techniques. There are 100 university graduate students. So according to this study, humor may be one of the best ways of teaching English as a foreign language.

     Algafar (2017) studied teachers’ perspectives toward the use of humor in teaching English as a foreign language. In this study, there are four English teachers that answer to seven questions that were asked by researchers. These questions are related to teachers’ ideas about advantages and disadvantages of using humor in the process of teaching in the classroom. Based on this study, teachers are agreed that there are so many advantages in use of collocations, such as sociological benefits, instructional benefits, and psychological benefits. Also, they refer to this point that one disadvantage of humor is that it is not acceptable by students in all situations.

     Piaw (2012) studied the effect of content-based humorous cartoons in teaching and learning materials to improve students’ reading rate, comprehension and motivation. In this study, there were 80 students with the same educational and background knowledge. They have selection test and tests for investigation of their affective, cognitive and communicative skills. Piaw conclude that in a group which was used content-based humorous cartoons in the course material were improved in their reading rate, comprehension, and reading motivation.

Ghanei Motlagh, Motallebzade, and Fatemi (2014) had studied the effects of teacher’s sense of humor on Iranian’s EFL learners’ reading comprehension ability. They had 58 students and divided them into two groups which received the same texts and process of teaching. The experimental group received the same texts and the same process of teaching but with humorous environment that managed by teacher. The other group had the same condition but without the humorous environment. So, they conclude that if teachers use humor in their classes, it is effective in improving learners' reading comprehension ability and enhancing intrinsic motivation.

Collocations

Collocations are universal linguistic phenomenon, and it is not logical to say that a natural language is free of collocations. Collocations are known as an important factor in any language. Therefore, English learners should pay attention to collocations. Most of the students try to memorize words independent of the context. Therefore, they feel that they know the meaning of every word but they cannot make sense of it. According to Wray (2002), there are different important functions of collocations. Firstly, collocations are very effective in language learning. It plays an important role in development of first and second language acquisition. Secondly, collocations are very important in fluency of spoken and written language. Thirdly, collocations have positive effect on the comprehension skill of students because they understand the meaning of a text without attending to all the words. The final function of collocations is that they are used as members of a linguistic group.

     There are several studies which refer to the teaching/ learning of collocations and the use of humorous texts in education. For instance, Rahimi and Momeni (2012) studied the effect of teaching collocations on English language proficiency. They believed that learning a second language involves four main skills which are known as elements of effective communication.Therefore, an important factor for this is the amount of vocabulary. They had sixty participants in pre/posttest quasi-experimental design study. The control group learned the vocabularies in isolated forms, but the experimental group were taught collocations of some particular words. They concluded that the experimental group outperformed the control group in the posttest, implying that teaching collocations can improve students’ language proficiency.

     Woolard (2000) studied the effect of teaching collocations on the reading comprehension of some advanced EFL students. The purpose of this study was to compare the effect of explicit collocation teaching and explicit single-item vocabulary instruction on the reading comprehension skill of the participants. One experimental group received explicit collocation teaching, and the other received single-item vocabulary instruction with reading passage. The findings showed significant difference across the reading comprehension of the two groups.

     Lewis (2000) studied the effect of implicit and explicit instruction of collocations on Iranian EFL high school students’ recognition. The purpose of this study was to investigate the effect of these two instructions on recognizing collocations of delexicalized verbs such as “make, do, give, take, and have” by EFL students. There were sixty students in this study, and they were divided into experimental and control groups. The first experimental group learnt collocations implicitly, and the other one learnt it explicitly. The first group learnt collocations through reading a text and then answering comprehension questions, and the second one read a text and did focus-on-form tasks. The findings indicated that the students who had learnt the collocations explicitly outperformed the other group.

     Kiaee, Moghaddam, and Moheb Hosseini (2013) examined the effect of teaching collocations on enhancing Iranian EFL learners’ reading comprehension. There were 70 intermediate EFL learners divided into two groups of experimental and control group. The results indicated that the experimental group was better than the control group. Therefore, collocations were shown to have a positive effect on the improvement of learners’ reading comprehension.

     Kim and Bae (2012) studied the relationship of collocation competence with reading and writing skills. There were 86 Korean students who read a text containing lexical and grammatical collocations. Then, they were given a test containing these collocations. Next, they were given a written test related to the content of the reading texts. The researchers concluded that there was not any relationship between students’ knowledge of collocations and their reading skill, but there is a relationship between knowledge of collocations and their writing ability.

 Method

Participants

Participants of this study were male and female intermediate EFL learners in Iran. At the beginning of this study, the number of the participants was 64, but after taking Success Test, proficiency test, only 59 of them remained and 5 of them were excluded from the study. 21 of them were male and 38 students were female. The number of male and female students was not equal because the equal number of male and female students was not available. Among these students, 42 of these students studied literature and others studied translation.

 Instruments

Success Test

The Matura success placement test was administered to determine the proficiency of EFL intermediate learners. It consists of 100 test items and the participants had 50 minutes to answer them.

Test of collocations

The second instrument was a collocation test containing 90 items that was used in order to determine the participants’ prior knowledge of collocation. The participants had 80 minutes to answer these items. The collocations were different from those which were used in the humor texts that were our purpose in the study. This test contains two parts: section one referred to as lexical collocation. This section comprised three parts: the first part included 15 multiple choice tests and the participants were asked to choose the correct answer. The second part incorporated 15 items which participants should write the correct answer in each blank; and the third part consisted of 15 items which were sentences in which one word was omitted and a Persian equivalent of the collocation in question was provided. In this part participants had 35 minutes to answer these questions. As with section two, it was designed in order to determine participants’ prior grammatical knowledge in English. This part contained of three parts; part one involved 15 multiple choice questions, part two entailed a sentence in which one word was omitted and participants should write the correct word which was suitable for previous word. Also, in this part participants had 35 minutes to answer these items. After measuring participants’ knowledge of collocation in order to know that whether participants know target collocations or not, we prepared a test which contained all collocations and students had to specify whether they know the collocations.

Data Collection Procedures

First, in order to determine the proficiency level of students, a Success Testwas given to the participants. Then, collocation test was given to them as the pretest to determine students’ collocation knowledge. The test entailed 25 items testing different kinds of collocations such as adverb + adjective (truly happy) and verb + noun (commit suicide) and verb + adverb (wave frantically). Then, the students received a humor text with particular collocations throughout five sessions. After reading the texts, the researcher collected humorous texts, and gave collocation tests to the students.

     In these tests in addition to collocation items, there were some vocabulary items that participants had to answer. In test one; there were 18 collocations with some vocabulary items. In the second test, in addition to 11 collocation items with some vocabulary items’, there were 18 collocations that were in test one. In the third test, there were 10 collocations plus collocations of test one and two. In the fourth test; there were 9 new collocations and previous collocations. In the fifth test, there were 11 new collocations plus previous collocations. And in final test, all collocations that were used in previous tests were repeated as a post-test. Table 1 shows all details related to the tests.

 Table 1

Detailed information of tests

As Table 1 shows, after reading text 1, students received a test which contained 23 items (18 collocation items and 5 vocabulary ones). In this test participants chose and found the correct answer in each blank. In test 2, there were 32 items which included 11 new collocations with 3 vocabulary items. In test 3, there were 44 items which had 10 collocation items with 5 vocabulary items. Test 4 had 54 items, nine of which were new collocation items. In test 5, there were 65 items which had 11 new collocations with 5 vocabulary items. In the post-test, without reading humorous texts, students received a test which contained 60 items that were used in the previous texts.

 Table 2

List of Collocations from Five Humor Texts

      As with the texts used in this study, there are five humor texts which have 60 collocations, totally. In each session participants received a humor text with particular collocations. The texts were selected from the famous book titled “Collocation in Use” the intermediate level. The texts were three to four paragraphs long. In the first session, participants received two short humorous texts. The first one contained 260 words with nine collocations. The other text contained 238 words again with nine collocations. In the second session, participants received two humorous texts that the first one contained 209 words and 5 collocations and the other text contained 208 words and six collocations. In the third session, they received three texts that the first one had 224 words with four collocations the second text had 97words with two collocations and the last text had 162 words with five collocations. In the other session, there were two short humor stories that the first one contained 434 words with 10 collocations and the other part contained 220 words with three collocations. In the final session, there were two humor texts, the first one had 218 words with seven collocations and the other one had 165 words with nine collocations.

     Table 3 shows the number of words, number of collocations, and readability for each text. Readability of each text was calculated with “Readability Formulas-Flesch Reading Ease score”. Based on this formula, most of texts were easy to read but our purpose was on the effect of humor texts on learning of collocations by intermediate students. So, their readability was not so important in this study.

Table 3

Details of Texts

  Data analysis

The statistical analyses of this study were done through using the 22^th version of IMB Statistical Package for Social Science (SPSS 22).The descriptive statistics of gender, field of study, Success Test, Collocation Test, pre and post-test were calculated.  In order to investigate the effect of humor text on learning collocations by EFL learners, the data of five tests were categorized in pairs in order to compare them in paired-samples t-test.  Also in this study, we analyzed the results with ANOVA. In order to find the effect of humor texts on participants’ knowledge of collocations, in this study we compared the results of pre and post-tests.

Results

Analysis of the Pre-Test

Table 4 below presents descriptive statistics for pre-test which is used in order to determine students’ knowledge of specific collocations used in humor texts.

     Based on Table 4, there are 59 students. The pre-test has a mean value of 36.61 with an SD value of 15.53. The lowest score in this test is 0 while the highest score is 59.33 out of 100.  The skewness and kurtosis value is -0.879 and .09, which show that the distribution of scores is normal.

 Paired-Sample T-Test

     In order to investigate the effect of humor text on learning collocations by male and female students, the researcher categorized each test in pairs in order to compare them in paired-samples t-test. Therefore, in this study, there were four pairs of scores which underwent paired-samples t-test statistical operations. The results of test one and two are in a pair, test two and three in the other pair, test three and four in the other, and finally test four and five in a pair. Table 5 presents the mean and standard deviation for the four pairs of scores.

 As shown in Table 5, the mean value of test 1 and 2 is 31.45 and 56.45 with SD value of 23.78 and 17.2. As is evident from the corresponding mean values, the only pairs where there was an increase in mean values from an earlier treatment to the next are pairs 1 and 4. For pairs 2and 3 there was a decrease in the mean value from an earlier treatment to the next which is not what was expected from the design of the study. This relationship is further investigated when the relevant t-test values are computed and statistically checked for their significance in the sections that follow. Table 4.6 presents the amount of correlation between the members of pairs. All pairs show statistically significant correlations except for Pair 2 where the obtained correlation coefficient is not statistically significant at p< .05.

Table 7 presents the results of paired-samples t-test calculations for the five pairs of this study. The first column shows the pairs and the second column presents the mean difference values between the respective pair members.

     Table 7, the negative values here mean that there was a decrease in scores from the earlier to the next treatment which was not what was expected based on the research design and research questions. Pairs 2 and 3 are in this situation. Pairs 1 and 4 show increases in the amount of their corresponding mean values from an earlier to the next treatment. Pair 1 shows the highest mean difference value of 24.53 with an SD value of 23.08. The amount of difference for pair 4 is not as large as that of pair 1.

The difference in the Total Collocation Score from the third treatment to the fourth one of pair 4 is only 3.22 with an SD value of 20.20. So far, the highest mean difference belongs to Pair one. What is even more important here is to investigate to see whether these observed differences in the mean values for the intended pairs in this study are statistically significant or not.

     Table 7, pair one and two are pairs whose observed difference between the mean values of its members is statistically significant at p

Analysis of the Post-Test

Descriptive statistics

     In order to find the effect of reading humor texts on learning of collocation, there was a post-test which contained all collocations used in five texts and prepared for students. Descriptive statistics of post-test are shown in Table 8.

     Based on Table 8, the mean value of post-test is 45.36 with standard deviation of 21.009. The skewness and kurtosis values in Table 4.8 show that the distribution of scores for both variables is normal because the relevant values reported in this table are very close to zero. Table 4.9 compares the results of post-test with each test separately.

    Table 9

    Comparison of Post-Test and other Tests

     As shown in Table 9, the mean value of post-test is 45.36 with standard deviation of 21.009.  In this table, there is a comparison between post-test and different tests of this study. The mean value of post-test is higher than Test, this shows that participants perform better in post-test rather than Test 1. But the mean value of post-test is lower than other tests.it is possible to conclude that participants were better in post-test rather than Test1 because they might learn collocations due to repetition. But they were not so good in others test in comparison with post-test. The mean value of post-test was 45.36 with standard deviation of 21.009  and the mean value  and standard deviation of test2 was 56.45 with standard deviation of 17.20, test3 was 46.32 with standard deviation of 18.81, test4 was 46.01with standard deviation of 20.72, test5 was 49.23 with standard deviation of 15.06. The performance of participants was better in test 2, 3, 4, and 5 rather than post-test. This might refer to the number of items which increase in post-test.

According to Bachman (1990) there are some factors which effects on participants performance such as testing environment, test rubric, etc. The first one refers to testing environment which includes: place of testing which are not the same in all tests of this study but participants were familiar with all of classes that had a test. The other factor is the personnel involved in the test that are different in each test. The final case refers to the time of testing.  Test 3 and 4 were in the afternoon but others in the morning. So, their performance might change based on the time of testing. They might better in the morning rather than in the afternoon.

     The other case refers to test rubric which includes: “ test organization: salience of parts, sequence of parts, and relative importance of part-, time allocation, and instructions-language, channel, specification of procedures and tasks, and explicitness of criteria for correctness. The first one is test organization that was the same in all tests. The other one is time allocation. In this study, there were five tests as treatment tests. In each test-based on the number of collocations- the number of items changed. But the amount of time allocated for each test was due to the number of items in each test and they have enough time for answering to each item. –one minutes for each test. In order to distinguish enough time for all of items of different test, we prepared pilot test. The other case is instruction such as language, channel (aural or visual), specification of procedures and tasks, and explicitness of criteria for correctness. In this case all of these factors are the same in tests

     In Table 10, there is comparison between post-test and each pairs of this study which is prepared in order to find the effect of reading humor texts on learning collocations.

 Table 10

Comparison of Post-Test and Four Paired Sample T-Test

      As shown in Table 10, there is a significant difference between the results of post-test and each pairs of this study. The mean value and standard deviation of post-test is 45.36 and 21.00 respectively. And the mean value of each pair with their standard deviation is (pair 1:24.53/23.08), (pair 2: 10.13/23.18), (pair 3: .31/23.88), and (pair 4: 3.22/20.20). there are so many factors which effect on participants performance and one of this factors may refer to the positive effect of reading humor texts on learning collocations. Based on Table 4.13, it is clear that participants was better than in post-test rather than other pairs of this study.

     One of question of this study refers to the difference between male and female performance in this study. So, in Table 4.11, there is a comparison between male and female’s performance in post-test and each pairs of this study.

Table 11

Comparison of Male and Female In Post-Test And Four Paired Samples

      Based on the results of Table 11, it is clear that female participants were better than male participants in this study. The mean value and standard deviation of post-test and four pairs of male participants are: (post-test: 45.36/21.00), (pair 1: 15.12/21.43), (pair 2: .39/20.51), (pair 3: 1.65/17.68), and (pair 4: 3.37/15.24). The mean value and standard deviation of post-test and four pairs of female participants are: (post-test: 45.36/21.00), (pair 1: 29.73/22.56), (pair 2: 15.51/23.06), (pair 3: .42/26.89), and (pair 4: 6.86/21.82).

Analysis of Pre and Post-Test

Descriptive statistics

In order to know whether reading humor texts had any effect on learning collocations or not, this part refers to the comparison between pre and post-test results. Table 12 refers to paired sample t-test of pre and post-test scores.

     Based on Table 11, the mean value of pre-test is 36.61with standard deviation of 15.53 and post-test has mean value of 45.36 with standard deviation of 21.009. So, it is clear that students’ scores in pre-test are lower than post-test. We can conclude that reading humor texts has positive effects in learning collocations by intermediate students.

Table 13 refers to analysis of pre and post-test score in order to find whether reading humor texts has any effects in learning collocations or not.

 Table 13

 Based on Pallent (2013), if the Sig. (2-tailed) is less than .05, we can conclude that there is a significant difference between two sets of scores in this study. Based on Table 13, the Sig. value is .019, so there is significant difference between these two sets of scores in this study.

 Analysis of ANOVA Results

Table 14 below presents the factors (in this case Time periods) used in this statistics. Five factors or five different time periods were used representing the five levels of the dependent variable of the study, namely the Test1, Test2, Test3, Test4, and Test5 during which data collection efforts were conducted and mean values on the continuous variable Total Collocation were obtained. 

      Table 15 below shows the mean and standard deviation values for the five levels of the dependent variable. The highest and lowest mean values are equivalent to 56.45 and 31.92 with SD values of 17.2 and 23.78 which belong to the first and the second time periods respectively.

      Table 16 below presents the results of multivariate tests as part of one-way repeated-measures ANOVA. This statistic was used due to the fact that the same participants are employed in different data collection efforts. The results of the one-way repeated-measures ANOVA are presented as follows:

     A one-way repeated-measures ANOVA test was conducted to compare the participants’ scores on collocation tests at Time 1 (during the treatment), Time 2 (during the treatment), Time 3 (during the treatment), Time 4 (during the treatment), and Time 5 (during the treatment). A significant effect for time was observed, Wilks’ Lambda = .44, F (5, 54) = 13.88, p < .0005, multivariate partial eta squared = .56. The effect size reported here as the multivariate partial eta squared value is indicative of a very large effect size due to guidelines proposed by Cohen (1988, pp. 284-7).

 Table 16

     The results presented above show that time period was a statistically influential factor but they do not tell us whether this change was favorable or not. In order to investigate whether the changes caused by the various time periods in the participants’ total collocation scores, Table 17 should be analyzed carefully. The differences in the mean values (statistically significant mean differences) between the members of the five pairs are indicative of the favorableness of the changes. The first row of the following table which shows the differences between the first time (Time 1) and the other 4 times. All of the mean differences are favorable but only the first four are statistically significant at p < .005. This suggests that the interventions (using humor texts) were effective in increasing the participants’ collocation knowledge as the mean differences show increases from time 1 (during intervention) to all the other times. The negative mean differences suggest that the mean values were higher in Times 2, 3, 4, 5compared to that of the time 1 (before intervention).  The second row (row 2) shows the mean differences between the second time (Time 2) and the other times. Except for the first mean difference value (from time 1 to time 2) which is 24.53 and is statistically significant as discussed above, the other mean values are not favorable as they show that the mean values have decreased from the second time period (Time 2) to all the other following Times due to the fact that the mean differences are positive values meaning that Time 2 had the highest mean value among all the other levels of the dependent variable. The third row (row 3) presents the mean differences between Time 3 and all the other Time periods. The statistically significant and favorable mean difference is observed between Time 3 and Time 1 which is equal to 14.40 and is statistically significant at p < .005.

     The other mean difference values in this row are either not statistically significant or not favorable or both. In row 4 the only statistically significant and favorable mean difference is the one between time 1 and time 4. The mean differences between time 2 and time 4 is -10.444 indicating that the participants’ collocation scores have actually decreased from time 2 to time 4 which is statistically significant but not favorable. In Row 5 the only statistically significant (p< .0005) and favorable mean difference is that from Time 1 to Time 5, meaning that the collocations mean scores have decreased from an earlier time to the next.

 Discussion

The current study aimed to observe the effect of reading humor texts on learning collocations by Iranian EFL learners. Based on the results, it was found that reading humor texts has positive effects on learning collocations by Iranian intermediate learners.

     According to Bachman (1990), there are three types of factors that affect students’ performance. These factors are test method facets, personal attributes, and random factors. The first one is systematic. This is a specific form of a test administration and it does not change based on the time of administration, or the purpose of the test. The second factor is students’ attributes that is not the purpose of a study. This refers to students’ cognitive style, their knowledge, and other characteristics such as sex, race, and ethnic background. This factor is systematic, too. The other factor which affects students’ performance is random factor that is unsystematic. This case refers to some features that are not predictable such as mental alertness or emotional state, and changes test method facets such as different test environment. So, based on these features the variation on participants’ performance may refer to these factors.

     The findings of this study are in line with Zabidin (2015), Algafar (2017), Atir (2010), and others who refer to this point that the use of humor has positive effect on learning. According to Hayati et al (2011), humor texts have positive effects on the process of learning. In their study whose purpose was to consider the effect of humorous texts in improving reading comprehension of EFL learners, humorous group improved in the recall and comprehension of the text. So, based on some factors that effects on participants’ performance, humor texts can help students in order to learn collocations. According to studies that are mentioned in Chapter Two, humor texts have positive effects on participants learning.

 Conclusion

The purpose of this study was to investigate the effect of reading humor texts on learning collocations by Iranian students. The results showed that reading humor texts has positive effects on learning collocations by EFL learners. As with the implications of the study, teachers should pay attention to the teaching of collocations and find different ways in order to help their students in order to improve their knowledge of collocations in order to get better in their different skills. These skills refer to their reading, speaking, listening, and writing abilities. Also, creative teachers, in order to create an environment which is suitable for their learners should decrease the pressure and anxiety of their classes. So, they can use humor in the process of teaching.

     One of the limitations of this study refers to a limitation in selection of students. There is not possibility to choose more intermediate students that were ready to participate in this study. The other one refer to the limitation in time of testing. The other limitation refers to limited number of collocations. If we want to choose more collocations, it is possible that we observe negative results because students have not enough time and patient for answering to all of questions.

     Instead of selecting just intermediate participants, it is possible to choose two groups of participants with two level of proficiency and year of study in order to find the effect of proficiency in the process of research. Instead of choosing one sample of participants that all of them read humor texts, it is possible to choose two groups of students such as: experimental and control group. The first one receives a set of collocations in humor texts and the other group receives the same set of collocation in non-humor text.