View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/C8RP00071A (Paper) Chem. Educ. Res. Pract., 2019, 20, 160-174

Scientific process skills test development within the topic “Matter and its Nature” and the predictive effect of different variables on 7th and 8th grade students’ scientific process skill levels

Cemal Tosun
Faculty of Education, Department of Science Education, Bartın University, Bartın, Turkey. E-mail: cemaltosun22@gmail.com

Received 7th March 2018 , Accepted 5th September 2018

First published on 5th September 2018

Abstract

The aim of this study was to develop a Scientific Process Skills Test (SPST) in the context of “Matter and its Nature”. It was investigated whether there was a predictive effect of demographical features and participating out-of-school learning opportunities across the 7th and 8th grade students’ Scientific Process Skill (SPS) levels. Quantitative research methods were used in this study. Data were collected from 289 middle school students for the validity and reliability of the test. The SPST consisted of 30 items and entailed three sub-dimensions (“basic scientific process skills, causal scientific process skills and experimental scientific process skills). The reliability coefficient of the test was calculated using the KR-20 formula and was found to be 0.84. The SPST was applied to 472 middle school students with the aim of determining whether there is a predictive effect of gender, grade level, school location, parent's education level and participation in out-of-school learning opportunities across the 7th and 8th grade students’ SPS levels. Multiple regression analysis was conducted to determine the effects of demographical features and out-of-school learning opportunities on students’ SPS levels. As a result, it was determined that the grade level, gender or mother's education level were important predictor variables that affect middle school students’ SPS levels. It was also determined that out-of-school learning opportunities such as participating in science fairs, designing projects or reading scientific journals had an important predictive effect on students’ SPS levels.

1. Introduction

The term science literacy has had different meanings over time. There is no universally accepted definition of science literacy (Roberts, 2007). Science literacy is explained by two different approaches (Roberts, 2007). In one of the approaches, students understand the science concepts and they use these meanings in the solution of individual, social and cultural problems they will encounter (Bybee, 2015). The other is access to information through socio-scientific issues.

There is no consensus with the meaning and content of science literacy. Some definitions focus on facts, concepts and vocabulary, while other definitions emphasize scientific process skills (SPSs) and reasoning skills (Miller, 2006).

Science literacy has become a common vision of today's science education. Reforms are being carried out in this field in the United States, Canada, and Europe (Bybee, 1995; National Research Council (NRC), 1996). The science curriculum's vision in Turkey is to educate all students as science literate regardless of their individual differences (Ministry of National Education (MoNE), 2017). It is stated that individuals who are science literate should have an understanding of the nature of science, the basic concepts of science, the positive attitude, perception and value related to science, the interrelationships of science, engineering, technology, society, and the environment, and the scientific and technical psychomotor skills and SPSs (MoNE, 2017).

The primary goals of science literacy were emphasized by American Association for the Advancement of Science [AAAS] (1993) and the NRC (1996) as an improvement in thinking skills including computation, prediction, manipulation, observation, communication and critical response skills and the ability to engage in research activity including making observations, asking questions, planning research studies, reviewing what is known in the light of experimental evidence, using data collection tools, analyzing data, interpreting data, suggesting answers, and explaining, predicting and communicating results. These skills, which are the primary goals of science literacy, are among the SPS dimensions, and the SPS dimensions and the primary goals of science literacy intersect.

The PISA 2006 definition of science literacy has three dimensions: scientific concepts, scientific processes and scientific situations (Organisation for Economic Cooperation and Development (OECD), 2007a). These scientific processes are related to the recognition of scientific questions, the identification of evidence, the drawing of conclusions, the communication of these conclusions and the demonstration of understanding of scientific concepts (OECD, 2007a). The literacy skills of science students in the learning process can be improved as an alternative way through a process skills approach (Suryanti et al., 2018). Science literate individuals locate, collect, analyze, and evaluate the source of scientific and technological information. They use these resources to solve problems and make decisions (Holbrook and Rannikmae, 2009).

In order for each individual to be trained as a scientist, SPSs must be acquired and used individually. Individuals who acquire SPSs solve daily life problems with the ways, methods, and perspectives that scientists use.

It is not possible to realize the basic aims of the science curriculum without improving the skills of gathering, processing, and generating evidence from data. It is expressed in the literature that data processing skills have a central role in reasoning (NRC, 1996; Next Generation Science Standard (NGSS), 2013). Reasoning with data is important for conducting authentic inquiry in science and involves learning mathematical concepts and procedures and recording, interpreting, and representing data (Zimmerman, 2007).

Science content and SPSs are two primary components of science education. SPSs are very important in science teaching (Kujawinski, 1997; Myers et al., 2004). For many years it has been stated that the main purpose of science education is to gain SPSs (Germann, 1989; Germann et al., 1996). Besides, it is advised that students gain SPSs by means of several activities in science education (Huppert et al., 2002). According to NRC (1996), one of the aims of science education is to use SPSs when making decisions.

SPSs are basic skills that facilitate learning in science, enable students to be active, improve the sense of responsibility in their own learning, and increase the permanence of learning (Çepni et al., 1997, pp. 74–83). SPSs are classified as basic and high-level SPSs by some researchers considering the developmental periods of students (Saat, 2004; Bağcı-Kılıç, 2006; Rezba et al., 2007). Çepni et al. (1997), on the other hand, divide these skills into three as basic scientific process skills (B-SPS), causal scientific process skills (C-SPS), and experimental scientific process skills (E-SPS). The B-SPS dimension includes observing, classifying, measuring, communicating, and recording data sub-dimensions. The C-SPS dimension includes inferring, predicting, defining operationally and identifying variables sub-dimensions. The E-SPS dimension consists of sub-dimensions such as hypothesizing, designing an experiment, changing and controlling variables, modeling and data interpretation (conclusion-decision). Zimmerman (2007) classifies SPSs as specific for a field that states the necessity of knowing certain scientific concepts to be able to solve the problems related to a scientific field and general SPSs.

A literature review reveals that Scientific Process Skills Test (SPST) development studies have been carried out for half a century (e.g.:Tannenbaum, 1971; Molitor and George, 1976; Dillashaw and Okey, 1980; Tobin and Copie, 1982; Burns et al., 1985; Mckenzie and Padilla, 1986; Shahali and Halim, 2010). In terms of SPS studies in Turkey, the SPST developed by Burns et al. (1985) was adapted to Turkish by Geban et al. (1992), and this test was used in numerous studies in Turkey. In later years, several other test development studies for identifying SPSs were encountered (e.g.: Aydoğdu et al., 2012; Feyzioğlu et al., 2012).

A literature review reveals how students’ SPS level changes according to variables such as gender (e.g.: Zeidan and Jayosi, 2015; Güden and Timur, 2016), school location, (e.g.: Zeidan and Jayosi, 2015), grade level (e.g.:Böyük et al., 2011), parents’ education level (e.g.: Karar and Yenice, 2012a; Ocak and Tümer, 2014), family's income level (e.g.: Karar and Yenice, 2012a), the number of brothers/sisters (e.g.: Güden and Timur, 2016), having a computer (e.g.: Böyük et al., 2011; Ocak and Tümer, 2014), having a study room (e.g.: Böyük et al., 2011), school type (e.g.: Ocak and Tümer, 2014), profession of parents, a kindergarten school attended and the frequency of visit to natural areas (e.g.: Güden and Timur, 2016). It is stated in the educational research literature that there are many studies that have been conducted on some factors affecting students’ SPS level. However, it has been seen that these studies are carried out on only a few factors as independent variables. However, not all variables have been looked at in one study, and the degree of importance of some predictor variables affecting students’ SPS level was not examined.

Importance of the study

The current literature reveals that the SPST designed to assess basic skill levels is being applied to individuals with high levels of skill (Feyzioğlu et al., 2012). Moreover, it has been reported that implementing the same test in all fields and education levels and disregarding the knowledge of the field while developing test items causes controversy in terms of the validity and reliability of the test (Zimmerman, 2000, 2007; Feyzioğlu et al., 2012). In order to solve the problem related to a scientific field, it is necessary to know some scientific concepts in the related field (Zimmerman, 2000, 2007). For this reason, it is not right to evaluate students’ SPS in an unknown subject (Feyzioğlu et al., 2012). SPSs are used with content. It is important to know the content as well as the ability to use these skills. The content of the items prepared to measure the SPS of the students includes the evaluation of the understanding ability of the concepts (Harlen, 1999).

The curriculum of science teaching in Turkey emphasizes the training of science literate individuals. Training individuals with enhanced SPS is among the components of science literacy. To achieve this, taking precautions that will improve the students’ SPS from early ages is of importance. SPSs include skills that can be used at every stage of everyday life by being science literate by understanding the nature of science and thereby increasing the quality and standards of life (Williams et al., 2004). This brings about the necessity of performing test development studies descriptive of the SPSs of middle school students. To this end, in this study a SPST was developed for 7th and 8th grade students by taking the aforementioned drawbacks into consideration. Test development requires a long time period and expertise. The context of the test was based on “Matter and its Nature”, one of the four subjects in the science curriculum in Turkey. This topic includes the units: “Matter and Change of State”, “Matter and Heat”, “Pure Matter and Mixtures” and “Matter and Industry”. Also, the test was designed to be appropriate for 7th and 8th grade students so as to refrain from validity and reliability discussions caused by performing the test on all fields and grades, and disregarding field knowledge while developing test items.

Students’ science literacy is measured through socio-scientific situations in PISA (Programme for International Students Assessment) exams prepared for determining the level of comprehension, mathematics and science literacy by 15 year olds. It is based on the second approaches of measuring students’ science literacy in PISA exams (Bybee and McCrae, 2011). Three qualifications for science literacy in the PISA 2015 science framework text have been identified. These are: explain events scientifically, evaluate and design scientific investigations and interpret data and evidence scientifically. These three qualifications are measured in a socio-scientific context. It is influenced by the individual's knowledge and attitudes (OECD, 2013). The PISA 2006 results reveal that science education at the middle school (5–8 grades) level in the world is inadequate in achieving the targeted knowledge, skills and understanding (OECD, 2007b). In this study, the content of the items in the SPST was in the form of problem scenarios prepared from socio-scientific issues. With these problem scenarios, it was aimed to determine the level of SPS of 7th and 8th grade students. The reason of preference for 7th and 8th grade students was that they were the closest group to taking the PISA exam.

The science course consists of physics, chemistry, biology and earth science concepts in the middle schools in the Turkish curriculum. The topic “Matter and Its Nature” is in the chemistry field, and it has a wide coverage in the science curriculum in grades 5 to 8. Science teaching is usually carried out in three learning environments which are classrooms, laboratories and out-of-school learning environments (Orion and Hofstein, 1994). Out-of-school learning environments are based on the assumption that the learning as an individual activity can be performed in all areas of life not just in class. Practice areas are all kinds of teaching environments that will support formal learning. Science education in middle schools should expand beyond the classroom walls, and out-of-school learning environments offer many opportunities for students to learn science (Carrier, 2009). But the author did not find any research related to the predictive effects of participating in out-of-school learning opportunities (participating in science trips, designing projects, participating in science competitions, reading scientific journals and participating in science fairs) across students’ SPS level. In this study, whether the SPS level of middle school students changed according to participation in scientific activities that differ from the variables aforementioned was identified.

Many studies have found that the level of SPS of students is low (Walters and Soyibo, 2001). It is known that Turkish middle school students’ score is below the international average in PISA and TIMMS (Trends in International Mathematics and Science Study) exams in the field of science literacy (Gonzalez and Miles, 2001; Martin et al., 2008). Knowing what is most effective in improving students’ SPS is important to increase the skill level.

SPSs are acquired during the learning process. Hence the factors affecting SPSs are found in the process itself (Nwosu, 1991). Some factors that influence students’ acquisition of SPSs include the following (Martina, 2007): teacher's competence, the method of teaching, instructional materials, parents’ background and practice, gender, school location, school type, teacher sensitization, students’ cognitive ability, teachers’ experience, cognitive demand of students, and family's socio-economic status.

An investigation of the SPS levels of students and the variables affecting these skills is an important field of study in the process of deciding what needs to be taken into account in improving education. A literature review reveals how students’ SPS levels change according to some factors. However, it has been seen that these studies are carried out based on only a few factors as independent variables. However, there are no studies in the literature regarding the degree of importance of some predictor variables affecting students’ SPS level.

In this study, after confirming the validity and reliability of the developed test to identify the SPS level of middle school students, it was investigated whether there were predictive effects of demographical features (gender, grade level, school location, and parent's education level) and participating in out-of-school learning opportunities across the 7th and 8th grade students’ SPS levels with the analysis of the data collected from a large sample group.

Aims of the study

The aim of this study was two-fold: the first aim was to develop a Scientific Process Skills Test (SPST) within the context of “Matter and its Nature”. The second aim was to investigate whether there are predictive effects of demographical features (gender, grade level, school location, and parent's education level) and participating in out-of-school learning opportunities across the 7th and 8th grade students’ B-SPS, C-SPS, E-SPS and T-SPS level.

2. Method

The survey research design which is one of the quantitative research methods was used in the test development stage of this study. The survey research design is widely used in education as it is efficient and generalizable (Fraenkel and Wallen, 2003; McMillan and Schumacher, 2010). In survey research, the aim is to collect data to identify the specific characteristics of a group. Survey research design is often used for gathering and presenting data from large groups (McMillan and Schumacher, 2010). The non-experimental correlational-research design was used while identifying whether there were predictive effects of demographical features and participating in out-of-school learning opportunities across the 7th and 8th grade students’ SPS levels. In correlational-research design, it is examined whether there is a correlation between two or more groups or phenomena. The researcher determines the level of the relationship between two or more variables using statistical techniques. This design is preferred to see if the variables influence each other (McMillan and Schumacher, 2010).

Sample group

The data were collected from 289 students from six middle schools in Bartın in the northwest of Turkey for the validity and reliability of the SPST. The sample group for the test development stage is detailed in Table 1.
Table 1 The sample group for the development of the SPST
Frequency Percent (%)
Gender
Female 169 58.5
Male 104 35.9
Unspecified 16 5.6
Total 289 100
Grade level
7th grade 152 52.6
8th grade 137 47.4
Total 289 100

The test was applied to 472 middle school students in Bartın, Kilis, Antalya, Ordu, Kahramanmaraş and Konya in Turkey with the aim of determining whether there are predictive effects of demographical features and participating in out-of-school learning opportunities across the 7th and 8th grade students’ SPS levels. It was not claimed that the sample group represents all 7th and 8th grade students in Turkey. Sampling was selected from the students who study in different regions and have different economic status. Thus, it is provided that the sample group represents the properties of the population partly, though not completely. In addition, the number of observations was increased to allow the normal distribution of the data (Stevens, 2009, p. 221).

This study was carried out in the 2016–2017 academic years. To collect data from the students, necessary permissions were obtained from the school administrations. The SPST was conducted by teachers in the MoNE. The intended purpose of the test was expressed to the students. Students were asked to participate in the study. Participation in the test was based on volunteerism. They were told that they would not have any advantages or disadvantages of participating in the test. Detailed information on this sample group can be seen in Table 2.

Table 2 The sample group (sampling) for the application of the SPST (N = 472)
Frequency Percent (%)
Gender Female 213 45.1
Male 257 54.4
Unspecified 2 0.5
Grade level 7th grade 249 52.8
8th grade 223 47.2
School location Urban location (City center) 371 78.6
Rural location (Village) 93 19.7
Unspecified 8 1.7
Mother's education level Up to primary school (including illiterate and literate) 181 38.3
Middle school 115 24.3
High school 94 19.9
University 71 15.1
Unspecified 11 2.3
Father's education level Up to primary school (including illiterate and literate) 86 18.2
Middle school 87 18.4
High school 144 30.5
University 140 29.7
Unspecified 15 3.2
Scientific activities Science trips 116 24.5
Science fairs 190 40.2
Scientific journals 226 47.9
Science competitions 55 11.6
Design projects 168 35.5

Data collection tools

The test which was developed in the first stage of this study was used as the data collection tool in the second stage. In this stage, the acquisitions within the context of “Matter and its Nature” were examined. This topic includes the units: “Matter and Change of State” in the 5th grade, “Matter and Heat” in the 6th grade, “Pure Matter and Mixtures” in the 7th grade, and the “Matter and Industry” in the 8th grade. Within these units;

The topics of Change of State of Matter, Distinctive Features of Matter, Heat and Temperature and Heat Effects Matters are included in the 5th grade curriculum.

The topics of Particulate Nature of Matter, Condensation, Matter and Heat, and Fuels are included in the 6th grade curriculum.

The topics of Particulate Nature of Matter, Pure Matter, Mixtures, Decomposition of Mixtures and Domestic Waste and Recycling are included in the 7th grade curriculum.

Finally, in the 8th grade curriculum, the topics of Periodic System, Physical and Chemical Changes, Chemical Reactions, Acids and Bases, Matter–Heat Interaction and Chemical Industry in Turkey are included. What concepts these items include and which SPS levels they belong to are given in Table 3.

Table 3 The SPS levels of the items and the concepts contained in the test items
Scenarios Item Concepts SPS levels
Unmixed waters 1 Heat exchange, temperature, density and density difference Inferring
2 Space structure, moving structure, heat, heat exchange, temperature, density and density difference Conclusion-decision
3 Density and density unit Recording data
4 Density and temperature Communicating
Science laboratory 5 Evaporation and boiling Classifying
6 Heat, boiling, boiling points and temperature Predicting
Dry ice 7 Melting and sublimation Identifying variables
8 Melting and sublimation Identifying variables
9 Melting and sublimation Hypothesizing
Candle 10 Melting Identifying variables
11 Melting Identifying variables
12 Melting Changing variables
13 Melting Hypothesizing
14 Melting, domestic solid waste, recycling and reuse Defining operationally
Change of state 15 Evaporation, freezing, melting, deposition, sublimation and condensation Classifying
Snowman 16 Freezing point and melting Designing an experiment
Cold winter day in Erzurum 17 Freezing point Inferring
Emre playing ball 18 Evaporation, heat, heat exchange and temperature Conclusion-decision
19 Evaporation, heat, heat exchange and temperature Conclusion-decision
Temperature 20 Shrinkage, expansion, heat and temperature Classifying
Balloon tour 21 Expansion, heat and density Hypothesizing
22 Shrinkage, heat and density Inferring
23 Shrinkage, expansion and density Conclusion-decision
Bathroom 24 Heat exchange and temperature Measuring
Water and juice 25 Freezing point Hypothesizing
26 Freezing point Conclusion-decision
Circular metal 27 Heat conductivity Predicting
Oil 28 Solid fuels, liquid fuels, gas fuels, renewable and non-renewable energy fuels Inferring
29 Solid fuels, liquid fuels, gas fuels, renewable and non-renewable energy fuels Communicating
Cardboard 30 Heat conductivity Designing an experiment
Tea 31 Homogeneous mixture, solution, dissolution, factors affecting dissolution rate, contact surface, mixing, temperature, solvent and soluble Hypothesizing
32 Factors affecting the dissolution rate Predicting
Ayşe's experiment 33 Factors related to heat and core heat Predicting
Cabbage juice 34 Acid, base, indicator, mixture and Ph Conclusion-decision
35 Ph meter Measuring
Markets 36 Physical and chemical changes Classifying
Periodical system 37 Group, period, and classification of periodical system Observation
38 Nonmetal, metal and inert gas Observation
Imported and exported products 39 Imported and exported chemical products Observation
Cake 40 Acid, base, heterogeneous mixture and formation of chemical reactions Identifying variables
41 Acid, base, and formation of chemical reactions Identifying variables
42 Acid, base, and formation of chemical reactions Hypothesizing
Atom 43 Atom, the nature of scientific knowledge, nucleus and electron Modeling
44 Atom, the nature of scientific knowledge, nucleus, electron and layer Modeling

Test items were prepared in a way to include the aforementioned concepts at the middle school level. The items were included with B-SPS, C-SPS and E-SPS dimensions.

The content of the items in the SPST were in the form of problem scenarios prepared from socio-scientific issues by the researcher. In this stage, the types of questions, correct answers, distractors and what skill level is measured in scientific process skills tests in the related literature were examined. Some problem scenarios prepared in the science teaching and laboratory applications course which the researcher's carried out in previous periods were edited. Four academicians from different universities were consulted for opinion to ensure the suitability of the items. The opinions of five MoNE teachers who were attending courses at the Master's degree level in science education were also taken. The academicians and the teachers assessed the appropriateness of the items in terms of face and content validity. This way, the appropriateness of grade, concept and skill level was ensured. The skill levels of the test questions are determined by the researcher based on the types of questions in the literature. If there was hesitant as to whether some questions measure more than one skill level, and which skill level is more appropriate, these hesitations have been eliminated in the direction of the recommendations of the academicians and teachers. It is accepted that the questions measure a single skill level that is more concentrated. Thus, the final level of skill that the questions aimed at measuring was decided and Table 3 is prepared. Below, some of the items included in the B-SPS, C-SPS and E-SPS dimensions following the edition are given.

The B-SPS dimension sample item:

Item 15. Ayşe is a 5th grade student. Her teacher asks Ayşe to classify the concepts of melting, freezing, evaporation, condensation, deposition, and sublimation, which are in the “State Change of Matter” unit, into two groups according to heat exchange. Given that Ayşe performed this task correctly, which of the below do you think she must have done the classification with?

image file: c8rp00071a-u1.tif

The C-SPS dimension sample item:

On a hot summer day, Semih feels like having an ice cream after school and he goes to the ice cream shop. He buys his ice cream packed. The ice cream man asks him how long it takes to get home for he will put dry ice in the package (dry ice is solid carbon dioxide. It transforms from solid to gas, in other words, sublimation). Semih says the journey home will take an hour. Upon arriving home, Semih unwraps his ice cream to realize that the dry ice is not there. He thinks the dry ice has melted but he realizes that there is no water in the package.

Item 7. Which of the following is the independent variable in this scenario?

(A) The amount of dry ice

(B) That the ice cream did not melt

(C) The time it took to go home from the ice cream shop

(D) A hot summer day

The E-SPS dimension sample item:

Item 16. Tayfun starts watching cartoons after finishing his homework. In the cartoon, there is the father snowman, snow children and a rabbit. Father snowman takes the rabbit's carrots. The rabbit threatens to spill salt and melt his children to get his carrots back. Father snowman is very afraid… (Snow children are exactly the same in terms of height, weight, etc.)

Tayfun is very surprised by the fact that the rabbit can melt the snow children only with salt and without heating. To test this, he decides to design an experiment for the next day's science lesson. If you were Tayfun, what kind of experiment would you suggest him doing?

(A) He should put the same volume of ice in two identical containers and add different amounts of salt into each container. He should measure the melting point of the mixtures using a thermometer.

(B) He should put different volumes of ice in two identical containers and add different amounts of salt into each container. He should measure the melting point of the mixtures with a thermometer

(C) He should put different volumes of ice in two identical containers and add the same amount of salt into each container. He should measure the melting point of the mixtures using a thermometer.

(D) He should put the same volume of ice in two identical containers and add the same amount of salt into each container. He should measure the melting point of the mixtures using a thermometer.

The SPST was developed in Turkish. Data were collected from 7th and 8th grade students in Turkey. Test items were translated back into English by an English language expert. Test items were re-translated back into Turkish and the back translation was compared with the original (Taber, 2018). The science education expert examined the similarities of each item by comparing each item's original Turkish version and its translation back into Turkish. The science education expert rated the translation validity.

Data collection for SPST

In the test development stage of this study, data were collected at six middle schools in Bartın during the autumn semester of the 2016–2017 academic years. Then, SPST was administered to middle school students in six different provinces during the spring semester of the 2016–2017 academic years. The SPST was conducted by science and other branch teachers in the MoNE. Personal information was not collected from the participants. Also the students were given an hour for 30 questions in the SPST.

Data analysis

The students were given 1 point for each correct answer and 0 points for each wrong answer. The difficulty index, the discrimination index and the operating status of each item's distractor were identified during test development. The reliability coefficient of the test was calculated using the KR-20 formula. A confirmatory factor analysis was performed to set the sub-dimensions of the test. Also, the Cronbach's alpha reliability coefficient related to the sub-dimensions and the overall test was calculated. The Pearson correlation analysis technique was used to identify the inter-dimension relationship. In the next step, the data obtained from a large sample group was analyzed according to some predictor variables. At this step, as the data was distributed normally, multiple regression analysis was used.

3. Results

Results for aim 1

Test development process findings.
Item analysis findings. The answers given by 289 students to 44 questions in the test were analyzed with the item analysis method. The answer papers that were put through item analysis constitute a 27% portion of all the answer papers in the upper groups and lower groups. For this reason, item analysis was performed over the answer papers of 78 students in each group. In accordance with the data obtained, Table 4 was formed for the difficulty index, the discrimination index and the operational status of the distractor of each item.
Table 4 Difficulty and discrimination values of the items
Sub-dimensions Item Difficulty Discrimination Sub-dimensions Item Difficulty Discrimination
a An item which has appropriate item difficulty and discrimination values.
Observation 37 0.25 0.19 İnferring 1a 0.47 0.42
38a 0.37 0.38 17 0.46 0.29
39a 0.37 0.33 22a 0.40 0.36
Measuring 24 0.25 0.12 28 0.41 0.23
35a 0.56 0.51 Defining operationally 14a 0.56 0.53
Classifying 5a 0.65 0.46 Hypothesizing 9a 0.41 0.36
15a 0.61 0.58 13a 0.63 0.67
20a 0.49 0.46 21a 0.44 0.38
36 0.29 0.26 25a 0.39 0.35
Communicating 4a 0.39 0.42 31 0.24 0.18
29a 0.49 0.36 42a 0.44 0.31
Recording data 3a 0.53 0.58 Designing an experiment 16a 0.37 0.26
Predicting 6a 0.43 0.37 30 0.33 0.29
27 0.31 0.28 Modeling 43 0.47 0.27
32a 0.45 0.33 44a 0.39 0.29
33 0.41 0.28 Conclusion-decision 2 0.39 0.24
Identifying variables 7a 0.51 0.41 18 0.31 0.18
8a 0.44 0.50 19a 0.46 0.31
10a 0.51 0.56 23a 0.62 0.68
11a 0.29 0.44 26a 0.33 0.33
40a 0.41 0.33 34 0.45 0.26
41a 0.36 0.51 Changing variables 12a 0.35 0.19

As shown in Table 4, since the discrimination indexes of item 37 in the observation sub-dimension, item 24 in the measuring sub-dimension, item 36 in the classifying sub-dimension, items 27 and 33 in the predicting sub-dimensions, items 17 and 28 in the inferring sub-dimensions, item 31 in the hypothesizing sub-dimension, item 30 in the designing an experiment sub-dimension, item 43 in the modeling sub-dimension, and items 2, 18 and 34 in the conclusion-decision sub-dimension are below 0.30, these questions cannot be regarded as adequately distinctive (Ebel and Frisbie, 1986) and they were excluded from the test. On the other hand, item 16 in the designing an experiment sub-dimension, item 44 in the modeling sub-dimension and item 12 in the changing variables’ sub-dimension were kept in the test considering their distractors’ operational statuses and the fact that the test includes items related to these sub-dimensions.

Confirmatory factor analysis findings

Subsequent to item analysis, the initial 44 items were reduced to 31. Confirmatory factor analysis (CFA) was conducted to verify the accuracy of the data gathered in the primary stage of this study (Tabachnick and Fidell, 2001). CFA was conducted via the LISREL 8.8 statistics program. The t values of the 31 items in the B-SPS, C-SPS and E-SPS dimensions concerning the latent variables’ explanatory statuses of the observed variable and the error variance of the observed variables are given in Table 5.
Table 5 Confirmatory factor analysis results
Item no. B-SPS C-SPS E-SPS
t-Value Error variances t-Value Error variances t-Value Error variances
3 8.64 0.71
4 4.59 0.91
5 6.67 0.82
15 8.10 0.74
20 7.02 0.80
29 2.83 0.96
35 3.86 0.94
38 3.63 0.94
39 2.92 0.96
1 5.16 0.89
6 4.97 0.90
7 6.12 0.85
8 5.63 0.87
10 9.25 0.68
11 5.72 0.87
14 7.52 0.78
22 4.42 0.92
32 2.80 0.97
40 1.78 0.99
41 4.93 0.90
9 2.52 0.97
12 2.01 0.98
13 9.21 0.68
16 2.43 0.98
19 6.18 0.85
21 5.85 0.86
23 8.75 0.71
25 3.85 0.94
26 2.08 0.98
42 2.09 0.98
44 2.05 0.98

An examination of Table 5 shows that item 40 is not significant. It can be seen that items 9, 12, 16, 26, 42, and 44 are 0.05 significant while all remaining items are 0.01 significant. At this point, item 40 was excluded. While there are numerous fit statistics preferred in structural equation modeling, the results of fit statistics used frequently are presented in Table 6.

Table 6 Results of fit statistics
χ 2/sd GFI AGFI RMSEA CFI NNFI RMR SRMR
2.01 0.84 0.82 0.0059 0.84 0.83 0.017 0.070

In Table 6, the χ2/sd ratio is below 3, which means a good fit index (Kline, 2005). The fact that the RMSEA value is below 0.08 indicates an acceptable fit index (Jöreskog and Sörbom, 1993). The fact that the GFI and AGFI indexes are below 0.90 indicate that the fit index is poor (Hooper et al., 2008). The fact that the NNFI and CFI fit indexes are below 0.90 shows a poor fit index. RMR and standardized RMR values below 0.05 mean a good fit index (Brown, 2006) while these values being below 0.08 means acceptable fit indexes. It is seen that the results of the analysis give a good fit index RMR value and an acceptable fit index SRMR value. According to fit statistics, it can be considered that this test is a good model with all fit statistics (excluding GFI, AGFI, CFI, and NNFI) and that it is a valid test with its factor structures.

Reliability

The reliability coefficient of the 30 items was calculated with the KR-20 formula. The reliability coefficient of the test was calculated to be r = 0.84. The average item difficulty index of the test was found to be 0.47. The optimal difficulty level for four-choice items is about 0.62 (Kaplan and Saccuzzo, 1997). According to this assumption, the average difficulty values of the test were a little difficult from the optimal difficulty level. The Cronbach's alpha internal consistency coefficient of each dimension was also calculated. The internal consistency coefficient of the B-SPS sub-dimension is 0.71; the internal consistency coefficient of the C-SPS sub-dimension is 0.72; and the internal consistency coefficient of the E-SPS sub-dimension is 0.62. The Cronbach's alpha internal consistency coefficient for the overall test is 0.80.

Inter-dimensional consistency

The relationship between the sub-dimensions of the test was also examined. The results of the Pearson correlation analysis are given in Table 7.
Table 7 Correlation between the dimensions of the SPST
Dimensions B-SPS C-SPS E-SPS T-SPS
a Correlation was significant at the 0.01 level.
B-SPS 0.50a 0.54a 0.81a
C-SPS 0.51a 0.82a
E-SPS 0.82a
Total scientific process skills (T-SPS) 1

According to Table 7, there is a significant relationship of 0.01 between the sub-dimensions of the SPST.

Results for Aim 2

Descriptive findings for the test. In the second stage of this study, the predictive effects of demographic features and participating in out-of-school learning opportunities on the students’ SPS level were investigated. For this purpose, the normality of the data obtained from 472 students was checked according to the measures of central tendency (mode, median and mean) and the coefficients of skewness and kurtosis. It was found that the mode, median and the mean values of the overall test and each dimension were close. Besides, skewness and kurtosis coefficients are given in Table 8.
Table 8 Skewness and kurtosis coefficient
Sub-dimensions Skewness Kurtosis
B-SPS 0.422 −0.173
C-SPS 0.393 −0.301
E-SPS 0.354 −0.371
T-SPS 0.696 0.146

When Table 8 is examined, it is acknowledged that the data is within the normal distribution range as the skewness and kurtosis coefficients for the overall test and the sub-dimensions are between −1 and +1 (Morgan et al., 2004, p. 49).

Reliability for the SPST

The reliability coefficient was calculated in the second stage of this study using the data gathered from 472 middle school students. Cronbach's alpha for all dimensions of the SPST was calculated to be 0.79. The Cronbach's alpha of the B-SPS sub-dimension is 0.71; the Cronbach's alpha of the C-SPS sub-dimension is 0.64; and the Cronbach's alpha of the E-SPS sub-dimension is 0.67. The results show an acceptable reliability coefficient for an achievement test and reveal that the measurements are reliable (Shum et al., 2006).

Confirmatory factor analysis

CFA was conducted in the second stage of this study to verify the accuracy of data gathered from 472 middle school students. CFA was conducted via the LISREL 8.8 statistics program. The χ2/sd ratio (720.56/402 = 1.79) is below 3, which means a good fit index (Kline, 2005). The fact that the RMSEA value (0.041) is below 0.05 indicates a good fit index (Jöreskog and Sörbom, 1993). And the fact that the CFI (0.87), NNFI (0.86) and AGFI (0.89) indexes are below 0.90 indicates that the fit index is poor (Hooper et al., 2008). The fact that the GFI fit indexes (0.91) are above 0.90 shows an acceptable fit index. RMR and standardized RMR values below 0.05 mean good fit index (Brown, 2006) while these values being below 0.08 means an acceptable fit index. It is seen that the result of the analysis gives a good fit index RMR value (0.012) and an acceptable fit index SRMR value (0.052).

Multiple regressions

All of the predictive variables were nominal. Predictive variables such as gender, grade level, school location, participating in science trips, designing projects, participating in science competitions, reading scientific journals and participating in science fairs are the qualitative variables with two categories. Before the regression analysis, gender was coded as “male = 0” and “female = 1”. Grade level codes were determined as “7th grade = 0” and “8th grade = 1”. School location codes were determined as “rural location = 0” and “urban location = 1”. Science trip codes were determined as “participating in science trips = 1” and “not participating in science trips = 0”. Project codes were determined as “designing projects = 1” and “not designing projects = 0”. Science competition codes were determined as “participating in science competitions = 1” and “not participating in science competitions = 0”. Scientific journal codes were determined as “reading scientific journals = 1” and “not reading scientific journals = 0”. Science fair codes were determined as “participating in science fairs = 1” and “not participating in science fairs = 0”. Mother's education level and father's education level had more than two categories and they were transformed into two categories. Thus two categorical dummy variables were obtained. The first category is the primary school and middle school education level. The second category is the high school and university education level. Mother's education level and father's education level codes were determined as “high school and university education level = 1” and “primary school and middle school education level = 0”. In multiple regression analysis, all of the predictive variables were included in the model. This approach is called the standard method (Enter method). The purpose is to examine the effect of all predictive variables on the predicted variable. Predictive variables must be independent of each other in order for the regression analysis to give correct results. That is, the predictive variables should not have a high degree of correlation among themselves. For this purpose, the correlations among the variables were examined and the results are presented in Table 9.
Table 9 Correlation between variables
Gender Grade level School location MEL FEL Trips Project Fair Competitions Journal
Mother's education level = MEL; father's education level = FEL.
Gender — 0.009 0.078 −0.066 0.020 0.046 0.155 0.033 −0.032 0.117
Grade level −0.028 −0.003 0.007 0.046 −0.110 0.197 0.017 0.038
School location 0.220 0.281 0.026 0.026 0.189 −0.051 −0.028
MEL 0.446 0.030 0.134 0.008 −0.009 0.137
FEL 0.085 0.123 0.123 0.049 0.127
Trips 0.093 0.013 0.159 0.036
Project −0.044 0.089 0.095
Fair 0.008 −0.058
Competitions −0.025
Journal —

There is a multi-correlation between variables with high correlations such as 0.80 to 0.90 (Field, 2005, p. 224). When Table 9 is examined, it is seen that there is no high correlation between the predictive variables.

The second aim is to investigate whether there are any predictive effects of demographic features (gender, grade level, school location, mothers’ education level and fathers’ education level) and participating in out-of-school learning opportunities (participating in science trips, designing projects, participating in science competitions, reading scientific journals and participating in science fairs) on the 7th and 8th grade students’ B-SPS, C-SPS, E-SPS and T-SPS level. Multiple linear regression analysis was conducted and the data obtained are given in Tables 10–13.

Table 10 Multiple linear regression analysis for the predictive effect of all variables on the B-SPS level
Variables B Std. error β t p Zero-order (r) Partial (r)
Constant 2.150 0.253 8.492 0.000
Gender 0.470 0.182 0.114 2.582 0.010 0.154 0.123
Grade level 0.798 0.181 0.194 4.398 0.000 0.192 0.206
School location 0.302 0.234 0.059 1.287 0.199 0.139 0.062
Mothers’ education level 0.675 0.209 0.158 3.227 0.001 0.249 0.153
Fathers’ education level 0.430 0.210 0.102 2.047 0.041 0.234 0.098
Trips −0.409 0.209 −0.085 −1.953 0.051 −0.032 −0.093
Project 0.726 0.190 0.170 3.821 0.000 0.205 0.180
Fair 0.330 0.187 0.079 1.764 0.078 0.129 0.084
Competitions 0.073 0.275 0.012 0.266 0.790 0.011 0.013
Journal 0.590 0.181 0.143 3.256 0.001 0.205 0.154

Table 11 Multiple linear regression analysis for the predictive effect of all variables on the C-SPS level
Variables B Std. error β t p Zero-order (r) Partial (r)
Constant 2.868 0.258 11.126 0.000
Gender 0.857 0.185 0.213 4.622 0.000 0.233 0.216
Grade level 0.414 0.185 0.104 2.241 0.026 0.120 0.107
School location −0.291 0.239 −0.059 −1.220 0.223 0.006 −0.058
Mothers’ education level 0.500 0.213 0.120 2.350 0.019 0.130 0.112
Fathers’ education level 0.076 0.214 0.019 0.356 0.722 0.100 0.017
Trips −0.174 0.213 −0.037 −0.814 0.416 0.001 −0.039
Project 0.449 0.194 0.108 2.322 0.021 0.151 0.111
Fair 0.457 0.191 0.113 2.397 0.017 0.121 0.114
Competitions 0.162 0.280 0.026 0.579 0.563 0.026 0.028
Journal 0.437 0.184 0.109 2.368 0.018 0.160 0.113

Table 12 Multiple linear regression analysis for the predictive effect of all variables on the E-SPS level
Variables B Std. error B t p Zero-order (r) Partial (r)
Constant 2.424 0.274 8.851 0.000
Gender 0.894 0.197 0.203 4.540 0.000 0.212 0.213
Grade level 0.719 0.196 0.163 3.663 0.000 0.167 0.173
School location 0.405 0.254 0.074 1.596 0.111 0.152 0.076
Mothers’ education level 0.790 0.226 0.173 3.495 0.001 0.228 0.165
Fathers’ education level 0.390 0.227 0.086 1.716 0.087 0.213 0.082
Trips −0.494 0.227 −0.096 −2.183 0.030 −0.033 −0.104
Project 0.384 0.206 0.084 1.866 0.063 0.138 0.089
Fair 0.309 0.202 0.069 1.528 0.127 0.128 0.073
Competitions 0.936 0.298 0.139 3.146 0.002 0.126 0.149
Journal 0.213 0.196 0.048 1.087 0.277 0.108 0.052

Table 13 Multiple linear regression analysis for the predictive effect of all variables on the T-SPS level
Variables B Std. error β t p Zero-order (r) Partial (r)
Constant 7.442 0.605 12.291 0.000
Gender 2.221 0.435 0.219 5.101 0.000 0.248 0.238
Grade level 1.931 0.434 0.192 4.449 0.000 0.199 0.209
School location 0.415 0.561 0.033 0.741 0.459 0.125 0.036
Mothers’ education level 1.965 0.500 0.188 3.931 0.000 0.253 0.185
Fathers’ education level 0.897 0.503 0.086 1.784 0.075 0.228 0.085
Trips −1.077 0.501 −0.092 −2.151 0.032 −0.027 −0.103
Project 1.560 0.455 0.149 3.430 0.001 0.204 0.162
Fair 1.096 0.448 0.107 2.449 0.015 0.156 0.117
Competitions 1.172 0.658 0.076 1.781 0.076 0.070 0.085
Journal 1.239 0.433 0.123 2.862 0.004 0.194 0.136

The predictive effect of all variables on the B-SPS level

Multiple linear regression analysis was conducted to identify the predictive effects of some demographical features (gender, grade level, school location, mothers’ education level, fathers’ education level) and participating in out-of-school learning opportunities (participating in science trips, designing projects, participating in science fairs, participating in science competitions and reading scientific journals) across students’ B-SPS level (see Table 10).

The results of the analysis reveal that (see Table 10) gender, grade level, school location, mothers’ education level, fathers’ education level, participating in science trips, designing projects, participating in science fairs, participating in science competitions and reading scientific journals show a statistically significant relationship (R = 0.451; R2 = 0.204) with students’ B-SPS level (F(10-435) = 11.123; p < 0.05). The 10 variables together explain 20.4% of the change in B-SPS scores. Grade level (β = 0.194, p < 0.05), designing projects (β = 0.170, p < 0.05), mothers’ education level (β = 0.158, p < 0.05), reading scientific journals (β = 0.143, p < 0.05), gender (β = 0.114, p < 0.05) and fathers’ education level (β = 0.102, p < 0.05) are significant predictors of the B-SPS level.

The predictive effect of all variables on the C-SPS level

Multiple linear regression analysis was conducted to identify the predictive effects of some demographical features and participating in out-of-school learning opportunities across students’ C-SPS level (see Table 11).

The results of the analysis reveal that gender, grade level, school location, mothers’ education level, fathers’ education level, participating in science trips, designing projects, participating in science fairs, participating in science competitions and reading scientific journals show a statistically significant relationship (R = 0.357; R2 = 0.128) with students’ C-SPS level (F(10-435) = 6.360; p < 0.05). The 10 variables together explain 12.8% of the change in C-SPS scores. Gender (β = 0.213, p < 0.05), mothers’ education level (β = 0.120, p < 0.05), participating in science fairs (β = 0.113, p < 0.05), reading scientific journals (β = 0.109, p < 0.05), designing projects (β = 0.108, p < 0.05) and grade level (β = 0.104, p < 0.05) are significant predictors of the C-SPS level.

The predictive effect of all variables on the E-SPS level

Multiple linear regression analysis was conducted to identify the predictive effect of gender, grade level, school location, mothers’ education level, fathers’ education level, participating in science trips, designing projects, participating in science fairs, participating in science competitions and reading scientific journals across students’ E-SPS level (see Table 12).

The results of the analysis reveal that gender, grade level, school location, mothers’ education level, fathers’ education level, participating in science trips, designing projects, participating in science fairs, participating in science competitions and reading scientific journals show a statistically significant relationship (R = 0.431; R2 = 0.186) with students’ E-SPS level (F(10-435) = 9.928; p < 0.05). The 10 variables together explain 18.6% of the change in E-SPS scores. Gender (β = 0.203, p < 0.05), mothers’ education level (β = 0.173, p < 0.05), grade level (β = 0.163, p < 0.05), participating in science competitions (β = 0.139, p < 0.05) and participating in science trips (β = −0.096, p < 0.05) are significant predictors of the E-SPS level.

The predictive effect of all variables on the T-SPS level

Multiple linear regression analysis was conducted to identify the predictive effects of some demographical features and participating in out-of-school learning opportunities across students’ T-SPS level (see Table 13).

The results of the analysis reveal that gender, grade level, school location, mothers’ education level, fathers’ education level, participating in science trips, designing projects, participating in science fairs, participating in science competitions and reading scientific journals show a statistically significant relationship (R = 0.493; R2 = 0.243) with students’ T-SPS level (F(10-435) = 13.932; p < 0.05). The 10 variables together explain 24.3% of the change in T-SPS scores. Gender (β = 0.219, p < 0.05), grade level (β = 0.192, p < 0.05), mothers’ education level (β = 0.188, p < 0.05), designing project (β = 0.149, p < 0.05), reading scientific journals (β = 0.123, p < 0.05), participating in science fairs (β = 0.107, p < 0.05) and participating in science trips (β = −0.092, p < 0.05) are significant predictors of T-SPS level.

When predictive effects of all variables were examined on middle school students’ SPS level, it is understood that the predictive effect on T-SPS scores (%24.3) is maximal and the predictive effect on C-SPS scores (%12.8) is minimal.

4. Conclusion and discussion

In this study, a scientific process skills test (SPST) within the learning topic “Matter and its Nature” appropriate for 7th and 8th grade students was developed. During this development process based on an item analysis and a confirmatory factor analysis, several items were excluded from the initial test and the finalized version of the test consisted of 30 items. The reliability coefficient of the test was found to be 0.84. The mean of the item difficulty index of the test was 0.47. The optimal difficulty level for four-choice items is about 0.62 (Kaplan and Saccuzzo, 1997). According to this assumption, the average difficulty values of the test were a little difficult from the optimal difficulty level. It has been reported in the literature that administering the same SPST on any field and any level of education, disregarding the field knowledge while developing the test items raised controversies in terms of reliability and validity (Feyzioğlu et al., 2012). In light of this, a SPST within the learning topic “Matter and its Nature” appropriate for the 7th and 8th grade students was developed and its validity and reliability were confirmed. The PISA 2006 results showed that the science education at the middle school level in the world was inadequate in achieving the targeted knowledge, skills and understanding (OECD, 2007b). Students’ science literacy is measured through social-scientific situations in PISA exams. In this study, the content of the items in the SPST were in the form of problem scenarios prepared from socio-scientific issues. Also, the test serves the purpose of being beneficial for teachers and researchers that aim at identifying the SPS level of students.

In the current study, the predictive effects of demographic features on the students’ B-SPS, C-SPS, E-SPS and T-SPS level was investigated. Within this aim, it was determined that gender in favor of female students is a significant predictor of B-SPS, C-SPS, E-SPS and T-SPS level. This finding is consistent with the results of Zeidan and Jayosi (2015)'s study which they conducted with middle school students, and reported that the SPS level of female students was higher on average than male students. Similar findings were reported by Karar and Yenice (2012a), who carried out a study with 8th grade students, and used the scientific process skills test that was developed by Burns et al. (1985). Conversely, it has been also reported in the literature that the SPS level of middle school students was not statistically different in terms of gender (Walters and Soyibo, 2001; Böyük et al., 2011; Güden and Timur, 2016).

In addition, the predictive effect of the grade level on the students’ B-SPS, C-SPS, E-SPS and T-SPS level was the other aim of the study. It was found that being at 8th grade was a statistically significant predictor of B-SPS, C-SPS, E-SPS and T-SPS level. In the study of Güden and Timur (2016), the researchers investigated the effect of grade level on the middle school students’ SPS, and reported that the SPS level of 5th, 6th and 8th grade students was higher than those of 7th grade students. Another study by Böyük et al. (2011) stated that there was a significant difference between 6th, 7th and 8th grade students’ SPS in favor of the 8th graders.

Another finding of this study was that school location did not have a predictive effect on the students’ B-SPS, C-SPS, E-SPS and T-SPS level. This result is not consistent with Zeidan and Jayosi (2015)'s study that was conducted with middle school students and reported that the SPS level of students in rural middle schools was higher on average than those in urban middle schools.

The results of this study revealed that mothers’ education level, in the case of the mother graduating from high school or university, was a significant predictor of students’ B-SPS, C-SPS, E-SPS and T-SPS level for 7th and 8th grade students. Böyük et al. (2011) also reported a significant difference between the SPS level of the students whose mothers are high school or university graduates and those of the students whose mothers are primary school graduates. A similar result was also put forward by Ocak and Tümer (2014). The studies by Karar and Yenice (2012a) and Germann (1994) also reported that the SPS of students differed according to their parents’ education level.

On the other hand, based on the results it was found that fathers' education level, in case of father graduated from high school or university, was a significant predictor of students' B-SPS level. The fathers’ education level did not have a predictive effect on the students’ C-SPS, E-SPS and T-SPS level. Böyük et al. (2011)'s study also expressed that there was a significant difference between the SPS level of the students whose fathers are university graduates and those of the students whose fathers are primary school graduates. Ocak and Tümer (2014)'s study reported similar results, stating that the SPS level of the students with university or high school graduate fathers was higher on average than those whose fathers are primary or middle school graduates.

Another question in this study was to investigate the predictive effects of participating in out-of-school learning opportunities (participating in science trips, designing projects, participating in science competitions, reading scientific journals and participating in science fairs) on the 7th and 8th grade students’ B-SPS, C-SPS, E-SPS and T-SPS level. The results of the study revealed that reading scientific journals or designing projects were significant predictors of B-SPS, C-SPS and T-SPS level. In addition, participating in science fairs was a significant predictor of C-SPS and T-SPS level. Karar and Yenice (2012b) suggested that various projects should be organized in order that the SPS level of the students be improved. This suggestion can be interpreted as the indicator of the difference between the SPS level of the students that design projects or not. It was reported that the improvement of students’ scientific discussion skills contributes to learning difficult topics in science centers (Guisasola et al., 2005, p. 549). It has been stated that trips to museums contribute to the improvement of the students’ observing, collecting data, inference and comment abilities (Chin, 2004; Griffin, 2004; Guisasola et al., 2005).

Based on the findings, the most important predictive variables on student's SPS level were gender, grade level and mothers’ education level from the demographical features when all variables were considered. The significant predictive effect of the grade level on the SPS level of the students can be interpreted as an indicator that any test cannot be applied to every age group. This result emphasizes the necessity of not ignoring the field knowledge while developing the skill test. From the studies conducted on the effect of gender on students’ acquisition of SPS, it was observed that the findings were inconsistent. While some studies reported a significant gender difference in the levels of SPS (Onukwo, 1995), others reported no significant difference in students’ acquisition of SPSs with respect to gender (Ofoegbu, 1984). Therefore, the effect of the gender variable on students’ level of SPS should be investigated in other studies in order to contribute to the debate. It has been also reported in the literature that the parents’ education background enhanced the rate of acquisition of the students’ SPS (Martina, 2007).

Additionally, the effects of out-of-school learning opportunities such as reading scientific journals, designing projects or participating in science fairs were important predictors of the students’ SPS level. The study of McMillan and Schumacher (2010) stated that scientific research was the process in which data are collected and the collected data are analysed in order to reach certain objectives. Thus, students learn the scientific research process when doing scientific research. The result of this study revealed that the students who read scientific journals, participate in science fairs and design projects were more engaged with SPS. Nnachi (1988) also reported that the students improve their SPS while visiting various interesting places (zoo, museum, botanical gardens, etc.) with their families. In Sullivan (2008)'s study, it was investigated how students utilized their thinking skills and SPS characteristics to solve robotics challenges. And, it was emphasized that learning robotics was related to thinking skills, SPS, and science literacy. The aforementioned study also expressed that the technological design and computer programming activities that are inherent in robotics projects as an out-of-school activity can provide an applied method of learning skills and SPSs.

It has been stated in the educational research literature that there are many studies which have been conducted on some factors affecting students’ acquisition of SPSs. However, these studies were only conducted with a few factors as independent variables. Also, the modeling studies that can explain the predictors of students’ SPS level in a holistic manner were very limited. In addition to the educational research literature, the degree of importance of some predictor variables affecting students’ SPS level has been determined in this study in a holistic manner.

5. Implications

The SPST developed in this study is a useful test for teachers and researchers that aim to determine the students' level of acquisition of SPS. The SPS levels of students are affected by several variables. But it is also important which of these is more effective. The degree of importance of some predictor variables affecting students’ SPS level was determined in this study. The present study revealed that the most important predictive variables on student's SPS level were gender, grade level and mothers’ education level from demographical features and reading scientific journals, designing projects and participating in science fairs from out-of-school learning opportunities.

Teachers should take students’ grade level and subject knowledge into account in the selection of data collection tools that will be used to determine students’ SPS levels. Because of the significant predictive effect of parents’ educational level on skills, family activities that will enhance the skill level of students are important. Basic skills are the prerequisite of higher level ones (Rambuda and Fraser, 2004). While basic skills can be earned from the pre-school period, high level skills can be gained from the middle school level. If parents are insufficient, this should be determined by teachers and these deficiencies should be accomplished by their teachers.

Schools should provide opportunities for students to do research. For this, teachers should encourage their students to do research, prepare a project and encourage them to participate in science fairs. Students should be offered learning environments to participate in such out-of-school learning opportunities.

Lavinghousez (1973) puts forward the idea that the optimum way of identifying the SPS level of individuals was through observations, presentations and laboratory practices. In order to be able to explain the change in the students’ SPS, it is necessary to show how much they understand and how much they can apply (Buck et al., 2008; Pyle, 2008). The reasons behind this situation can be detailed through interviews with students as well as through observation in their learning environment, which will reflect their skills. Educational policy-makers are interested about the factors that have the most effect on the development of students’ SPS. If policy-makers know what is most effective in improving students’ SPS, then would this help in their planning policies around these most influential factors. In this study, visiting science fairs from out-of-school learning activities was found to be among the most significant predictors of students’ SPS levels. Based on the findings, policies could be developed so as to include more extracurricular activities for students. Educational institutions are only responsible for the factors they can control. Therefore, for example, it is important to know which factors originate from institutions and which originate from students or other factors. There may be a predictive effect of other variables on students’ SPS level. The results of this study are limited to the predictive effects of demographic features and out-of-school learning opportunities on students’ SPS level. The predictive effect of the success level and the frequency of students’ using the laboratory, and affective factors such as attitude, perception and motivation on students’ SPS level can be investigated. It is difficult to develop a universal model that determines the predictors of SPS. More comprehensive studies on the predictors of SPS will be helpful for educational policy-makers, managers, researchers and teachers.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The author would like to thank the Scientific Research Projects Unit of Bartın University for financial support (Project Number 2016 SOS-A-005). I also thank CERP' editor and reviewers and Dr A. Çetin-Dindar, Dr E. Yılmaz, Dr H. Kaygın, Dr N. İlhan and Dr R. Yılmaz for their useful suggestions and discussions.

References

  1. American Association for the Advancement of Science (AAAS), (1993), Benchmark for science literacy. New York: Oxford Press.
  2. Aydoğdu B., Tatar N., Yıldız E. and Buldur S., (2012), The science process skills scale development for elementary school students, J. Theor. Educ. Sci., 5(3), 292–311.
  3. Bağcı-Kılıç G., (2006), İlköğretim bilim öğretimi [Primary science teaching], İstanbul: Morpa Kültür Yayınları.
  4. Böyük U., Tanık N. and Saraçoğlu S., (2011), Analysis of the scientific process skill levels of secondary school students based on different variables, J. TUBAV Sci., 4(1), 20–30.
  5. Brown T. A., (2006), Confirmatory factor analysis: for applied research, New York: Guilford.
  6. Buck L. B., Bretz S. L. and Towns M. H., (2008), Characterizing the level of inquiry in the undergraduate laboratory, Res. Teach., 38(1), 52–58.
  7. Burns J. C., Okey J. R. and Wise K. C., (1985), Development of an integrated process skill test, J. Res. Sci. Teach., 22(2), 169–177.
  8. Bybee R. W., (1995), Achieving scientific literacy, Sci. Teach., 62(7), 28–33.
  9. Bybee R. W., (2015), Scientific literacy, in Gunstone R. (ed.), Encyclopedia of science education, New York: Springer, pp. 944–947.
  10. Bybee R. and McCrae B., (2011), Scientific literacy and student attitudes: perspectives from PISA 2006 science, Int. J. Sci. Educ., 33(1), 7–26.
  11. Carrier S. J., (2009),The effects of outdoor science lessons with elementary school students on preservice teachers’ self-efficacy, J. Elem. Sci. Educ., 21(2), 35–48.
  12. Çepni S., Ayas A., Johnson D. and Turgut M. F., (1997), Fizik öğretimi: Milli eğitimi geliştirme projesi [Physics teaching: National education development project], Ankara.
  13. Chin C. C., (2004), Museum experience – a resource for science teacher education, Int. J. Sci. Math. Educ., 2(1), 63–90.
  14. Dillashaw F. G. and Okey J. R., (1980), Test of the integrated science process skills for secondary students, Sci. Educ., 64, 601–608.
  15. Ebel R. L. and Frisbie D. A., (1986), Essentials of educational measurement, Englewood Cliffs, NJ: Prentice-Hall.
  16. Feyzioğlu B., Demirdağ B., Akyıldız M. and Altun E., (2012), Developing a science process skills test for secondary students: validity and reliability study, Educ. Sci.: Theor. Pract., 12(3), 1887–1906.
  17. Field A., (2005), Discovering statistics using SPSS, London: Sage Publication.
  18. Fraenkel J. R. and Wallen N. E., (2003), How to design and evaluate research in education, 5th edn, London, UK.
  19. Geban Ö., Askar P. and Özkan G., (1992), Effects of computer simulations and problem-solving approaches on high school students, J. Educ. Res., 86, 5–10.
  20. Germann P. J., (1989), Directed-inquiry approach to learning science process skills: treatment effects and aptitude-treatment interactions, J. Res. Sci. Teach., 26(3), 237–250.
  21. Germann P. J., (1994), Testing a model of science process skills acquisition: an interaction with parents, education, preferred language, gender, science attitude, cognitive development, academic ability, and biology knowledge, J. Res. Sci. Teach., 31(7), 749–783.
  22. Germann P. J., Aram R. and Burke G., (1996), Identifying patterns and relationships among the responses of seventh-grade students to the science process skill of designing experiments, J. Res. Sci. Teach., 33(1), 79–99.
  23. Gonzalez E. J. and Miles J. A., (2001), TIMSS 1999 user guide for the international database: IEA's repeat of the third international mathematics and science study at the eighth grade, United States: Chestnut Hill, MA: The International Study Center, Boston College.
  24. Griffin J., (2004), Research on students and museums: looking more closely at the students in school groups, Sci. Educ., 88(1), 59–70.
  25. Güden C. and Timur B., (2016), Ortaokul öğrencilerinin bilimsel süreç becerilerinin incelenmesi (Çanakkale örneği) [Examining secondary school students’ cognitive process skills (Çanakkale sample)], Abant İzzet Baysal Üniversitesi Eğitim Fakültesi Dergisi, 16(1), 163–182.
  26. Guisasola J., Morentin M. and Zuza K., (2005), School visits to science museums and learning sciences: a complex relationship, Phys. Educ., 40(6), 544–549.
  27. Harlen W., (1999), Purposes and procedures for assessing science process skills, Assess. Educ., 6(1), 129–144.
  28. Holbrook J. and Rannikmae M., (2009), The meaning of scientific literacy, Int. J. Environ. Sci. Educ., 4(3), 275–288.
  29. Hooper D., Coughlan J. and Mullen M., (2008), Structural equation modeling: guidelines for determining model fit, Electron. J. Bus. Res. Meth., 7(2), 191–205.
  30. Huppert J., Lomask S. M. and Lazarowitz R., (2002), Computer simulations in the high school: students’ cognitive stages, science process skills and academic achievement in microbiology, Int. J. Sci. Educ., 24(8), 803–821.
  31. Jöreskog K. G. and Sörbom D., (1993), Lisrel 8: structural equation modeling with the SIMPLIS command language, Lincolnwood: Scientific Software International, Inc.
  32. Kaplan R. M. and Saccuzzo D. P., (1997), Psychological testing: Principles, applications, and issues, 4th edn, Pacific Grove, California: Brooks/Cole Publishing Company.
  33. Karar E. E. and Yenice N., (2012a), The investigation of scientific process skill level of elementary education 8th grade students in view of demographic features, Proc. – Soc. Behav. Sci., 46, 3885–3889.
  34. Karar E. E. and Yenice N., (2012b), Analysis of the scientific process skill levels of primary school 8th grade students based on different variables, Çukurova University Faculty of Education Journal, 21(1), 83–100.
  35. Kline R. B., (2005), Principles and practice of structural equation modeling, 2nd edn, NY: Guilford Publications, Inc.
  36. Kujawinski D. B., (1997), Assessment and evaluation of science process skills in secondary school biology laboratories, Doctoral Dissertation, Faculty of the Graduate School of State University of New York at Buffalo.
  37. Lavinghousez Jr. W. E., (1973), The analysis of the biology readiness scale (BRS), as a measure of inquiry skills required, Paper presented at BSCS Biology, College of Education, University of Central Florida.
  38. Martin M. O., Mullis I. V. S. and Foy P., (In collaboration with Olson, J.F., Erberber, E., Preuschoff, C., and Galia, J.), (2008), TIMSS 2007 international science report: findings from IEA's trends in international mathematics and science study at the fourth and eighth grades, United States: Chestnut Hill, MA: TIMSS & PIRLS International Study Center, Boston College.
  39. Martina E. N., (2007), Determinants of J.S.S. students’ level of acquisition of science process skills, Master Thesis, Nsukka: University of Nigeria.
  40. Mckenzie D. L. and Padilla M. J., (1986), The construction and validation of test of graphing in science, J. Res. Sci. Teach., 23(7), 1–9.
  41. McMillan J. H. and Schumacher S., (2010), Research in education: evidence-based inquiry, 7th edn, Boston: Pearson.
  42. Miller J., (2006), Civic scientific literacy in Europe and the United States, A paper presented to the annual meeting of the World Association for Public Opinion Research, Canada: Montreal, available online at http://www.arcsfoundation.org/Pittsburgh/JMiller.pdf.
  43. Ministry of National Education (MoNE), (2017), Ortaöğretim fen bilimleri dersi öğretim programı [Secondary science curriculum], Ankara.
  44. Molitor L. L. and George K. D., (1976), Development of a test of science process skills, J. Res. Sci. Teach., 13(5), 405–412.
  45. Morgan G. A., Leech N. L., Gloeckner G. W. and Barret K. C., (2004), SPSS for introductory statistics: use and interpretation, 2nd edn, London: Lawrence Erlbaum Associates.
  46. Myers B. E., Washburn S. G. and Dyer J. E., (2004), Assessing agriculture teachers’ capacity for teaching science integrated process skills, Journal of Southern Agricultural Education Research, 54(1), 74–85.
  47. National Research Council, (1996), National science education standards, Washington, DC: National Academy Press.
  48. NGSS Lead States, (2013), Next generation science standards: for states, by states, Washington, DC: The National Academies Press.
  49. Nnachi R. O., (1988), Acquisition of science process skills among junior secondary school students in Anambra state, Unpublished Master dissertation, Nsukka: University of Nigeria.
  50. Nwosu N. M., (1991), The effects of teacher sensitization on secondary school students’ level of acquisition of science process skills, Unpublished PhD dissertation, Nsukka: University of Nigeria.
  51. Ocak İ. and Tümer H., (2014), The level of having scientific process skills of 5th class primary school's students (Afyonkarahisar sample), Afyon Kocatepe University Journal of Science and Engineering, 14, 1–21.
  52. Ofoegbu L. I. J., (1984), Acquisition of science process skills among elementary pupils in some northern states of Nigeria, Unpublished PhD dissertation, Nsukka: University of Nigeria.
  53. Onukwo G. I. N., (1995), Development and validation of a test of science process skills in integrated science, Unpublished PhD dissertation, Nsukka: University of Nigeria.
  54. Organisation for Economic Co-operation and Development (OECD), (2007a), Assessing scientific, reading and mathematical literacy: a framework for PISA 2006, retrieved from http://www.oecd-ilibrary.org/education/assessingscientific-reading-and-mathematical-literacy_9789264026407-en.
  55. Organisation for Economic Co-operation and Development (OECD), (2007b), PISA 2006: Science competencies for tomorrow's world: volume 1-Analysis, retrieved from http://www.oecd.-ilibrary.org/education/pisa2006_9789264040014-en.
  56. Organisation for Economic Co-operation and Development (OECD), (2013), PISA 2015: Draft science framework, retrieved from https://www.oecd.org/pisa/pisaproducts/Draft%20PISA%202015%20Science%20Framework%20.pdf.
  57. Orion N. and Hofstein A., (1994), Factors that influence learning during a scientific field trip in a natural environment, J. Res. Sci. Teach., 31(10), 1097–1119.
  58. Pyle E. J., (2008), A model of inquiry for teaching earth science, Electron. J. Sci. Educ., 12(2), 1–19.
  59. Rambuda A. M. and Fraser W. J., (2004), Perceptions of teachers of the application of science process skills in the teaching of geography in secondary schools in the free state province, South Afr. J. Educ., 24(1), 10–17.
  60. Rezba R. J., Sprague C., McDonnough J. T. and Matkins J. J., (2007), Learning and assessing science process skills, Iowa: Kendall/Hunt Publishing Company.
  61. Roberts D. A., (2007), Scientific literacy/science literacy, in Abell S. K. and Lederman N. G. (ed.), Handbook of research on science education, Mahwah, NJ: Lawrence Erlbaum Associates, pp. 729–780.
  62. Saat R. M., (2004), The acquisition of integrated science process skills in a web-based learning environment, Res. Sci. Technol. Educ., 33(1), 23–40.
  63. Shahali E. H. M. and Halim L., (2010), Development and validation of a test of integrated science process skills, Proc. Soc. Behav. Sci., 9(2010), 142–146.
  64. Shum D., O'Gorman J. and Myors B., (2006), Psychological testing and assessment, Oxford: Oxford University Press.
  65. Stevens J. P., (2009), Applied multivariate statistics for the social sciences, 5th edn, New York: Taylor & Francis Groups.
  66. Sullivan F. R., (2008), Robotics and science literacy: thinking skills, science process skills and systems understanding, J. Res. Sci. Teach., 45(3), 373–394.
  67. Suryanti S., İbrahim M. and Lede N. S., (2018), Process skills approach to develop primary students’ scientific literacy: a case study with low achieving students on water cycle, IOP Conference Series: Materials Science and Engineering, 296(2018), 012030,  DOI:10.1088/1757-899X/296/1/012030.
  68. Tabachnick B. G. and Fidell L. S., (2001), Using multivariate statistics, Boston: Allyn and Bacon.
  69. Taber K. S., (2018), Lost and found in translation: guidelines for reporting research data in an ‘other’ language, Chem. Educ. Res. Pract., 19, 646–652.
  70. Tannenbaum R. S., (1971), Development of the test of science processes, J. Res. Sci. Teach., 8(2), 123–136.
  71. Tobin K. G. and Copie W., (1982), Development and validation of a group test of integrated science processes, J. Res. Sci. Teach., 19(2), 133–141.
  72. Walters Y. B. and Soyibo K., (2001), An analysis of high school student's performance on five integrated science process skills, Res. Sci. Techol. Educ., 19(2), 133–143.
  73. Williams W. M., Papierno P. B., Makel M. C. and Ceci S. J., (2004), Thinking like a scientist about real world problems: the Cornell institute for research on children science education program, J. Appl. Dev. Psychol., 25, 107–126.
  74. Zeidan A. H. and Jayosi M. R., (2015), Science process skills and attitudes toward science among Palestinian secondary school students, World J. Educ., 5(1), 13–24.
  75. Zimmerman C., (2000), The development of scientific reasoning skills, Dev. Rev., 20, 99–149.
  76. Zimmerman C., (2007), The development of scientific thinking skills in elementary and middle school, Dev. Rev., 27, 172–223.
This journal is © The Royal Society of Chemistry 2019