Stages of Marketing Research

Stages of Marketing Research

1. Identifying Problems and Opportunities

Research can be done to address the problem and find a solution, or to define new opportunities

Anthony Miles (BCG) formulated three key questions that must be answered at the stage of defining the problem or opportunity:

Why is the information it requested?

Was the information provided?

Can the question was really answered?

Capturing background

We must understand what the need to do some background research

Background information will determine the direction of the course of a research

There are three (3) general background researching links:

Want to make or modify products or services

Want to understand or predict behavior or market conditions

Want to affect some groups of customers => for example, along with the promotion or marketing communication program

Exploratory Research for Defining the Problem / Opportunity

Exploratory research is usually a small-scale research conducted to define the precise nature of the problem and to obtain a better understanding of the environment in which the problem occurred.

Methods commonly performed:

Focus Group Discussion

In-depth interviews (depth interview) or brainstorming with the experts

Analysis of secondary data

Case study

2. Determining Benefits and Interest Research

The end point of the process of formulation of the problem / opportunity is a statement of research objectives

Formulation good goal to act as a road map to develop research, and as a parameter in evaluating the quality of research

Objectives must be specific and clear

The purpose of research should avoid the syndrome of "nice to have"

3. Identifies Data Needs and Sources

Need to be identified first, whether to achieve the objectives of this research simply by using secondary data or should the primary data

Primary data: The results of the direct recording of the results of research, for example from interviews or questionnaires

Secondary Data: Data collected by the person or institution

Next specify the data source

Secondary Data

advantages:

Readily Available

Time, Cost, Power Less

Large scale

Some Information Only Available In Secondary Data

weaknesses:

May be less relevant

Data accuracy

Secondary Data Sources

Internal Data:

Sales Data

Data fees

Data Production

External Data:

Data Government Agencies

Data Syndication

Data Industry Association / Trade

The results of the study, the Internet, etc.

4. Choosing the type & Research Methods

Some research designs that can be selected are: surveys, observations, experiments

Selection of a research design was also influenced by the constraints of time and budget

Researchers must provide the best information to management by taking into account a variety of limitations

More detailed explanation of the research design will be discussed in the next module

5. Creating a Data Collection Instrument

For survey research, data collection instruments used were a questionnaire, interview or questionnaire. Structured questionnaire should be made to facilitate the interviewer to collect data from respondents.

For research observations and experiments, a tool used in general machinery observer (camera) and forms of observation / experiment

6. Designing Samples

Samples are elements that will be observed (respondent) which is a part of the population. Therefore, there are two questions that must be answered before selecting the sampling procedures, namely:

Who population?

Sample population: all people who use the aircraft through Soekarno Hatta airport in the last 1 year at least 3 times

Samples probabilistic or non-probabilistic?

Probabilistic Sampling Method

Simple Random Sampling

Suitable for the population as follows:

The number of members of the population is not too large

All members of the population has been registered

The condition is relatively homogeneous population

Stratified Random Sampling

Done by placing members of the population into several sub-populations or strata, then the sample drawn from each stratum separately so between strata being free stochastic.

Probabilistic Sampling Method

Systematic Random Sampling

Done by selecting a random sample and attract more examples at every distance k of sample units drawn previously.

K = N / n, where n = population size and n = number of samples to be taken

Cluster / Multi Stage Random Sampling

The elements of the population are grouped according to the location of the adjacent, then from the cluster randomly selected examples.

Sampling Method Non-Probabilistic

Purposive / Convenience Sampling

Based on the ease to obtain a sample is viewed from the side of the interviewer. Commonly used in exploratory research and customer intercept

judgment Sampling

The interviewer is testing its decision or experience in taking an example, with the intent of the results obtained will describe the target population

Sampling Method Non-Probabilistic

quota Sampling

This method is most excellent in the group nonprobability sampling.

Researchers conducted a strict control on the selection of examples which are based on one or more characteristics of the population, such as sex, age, education level, etc.

Objective: to get a proportional sample so as to describe the condition of the target population.

7. Collecting data

The main activity in the research activities is the collection of data

The data collection can be done by the existing units within the company or use a third party service (marketing research consultant)

Data is the raw material that will be processed in the research, so the quality must be good (valid, accurate) because  "Garbage in, Garbage Out"

Primary Data Collection Methods

For research surveys, in general there are two methods of data collection, the method of quantitative and qualitative methods

Quantitative Methods:

Personal Interview

House to House Interview

Office to Office Interview

Customer / Mall Intercept

Central Location Test

telephone Interview

Questionnaire by mail, email, mass media or at some point service

Qualitative methods

In-depth interviews (depth interview)

Focus Group Discussion (FGD)

For observational research, there is a quantitative method of data collection is often conducted primarily to assess the quality of services, namely Mystery Shopping

Mysteri Shopping

Is an observation that is commonly used to assess a service (banks, restaurants, hotels, airports, etc.)

Observer is a person who is not recognized by the parties that are observed, and behaves as a customer

Things were observed poured in a checklist form

These aspects are assessed generally consists of three aspects, namely the people, process, and tangible

Observer recruited by:

Experience conducting transactions on the object of observation

The ability of memory

Willing to make observations consistently

Observer attend training led by Project Manager / Field Coord.

Observer does not make a judgment on the observed object but only reveal the fact

Observer implement using observation agreed form (not filled at the place of observation)

Data Quality Control

Before the data is inputted and processed, first performed quality control activities to ensure that the data to be processed are valid

Activities quality control can be done in stages as follows:

Each questionnaire has been filled by officers will be checked by a Team Leader

Each questionnaire has been examined by the Team Leader will be examined by the Field Coordinator

Questionnaires that had escaped from the Field Coordinator then submitted to the Quality Controller

Quality Controller leading the back-check activities (revisit / recall) of 30% of the questionnaires at random

If the interviewer / observer cheating the whole questionnaire from the interviewer / observer will be dropped

8. Rework, Analyzing / Interpreting Data

Once the data is collected, the next step in the research process is to process and analyze the data. The purpose of this analysis is to interpret and draw conclusions from a number of data collected

The analysis technique used depends on the purpose of research, data collection methods, and the depth of analysis required

9. Delivering Reports

After processing and data analysis is complete, researchers must submit a report and submit conclusions and recommendations to management. This is a key step in the research process because the researcher must be able to ensure that research results can be trusted so that it can be input in the decision-making process

Follow-up Results of Research

After the company issued a number of energy and large sums of money to do the research, it is important to ensure that research findings can be used, followed up and used as a decision-making.

Evaluation Criterion: Validity

When the time to conduct research has come, we would deal with the following questions: Has the research method been accurate? Are the data we gained valid? Are questions that we used in the questionnaire reliable?

We must be able to answer those questions to ensure that our research study has followed the principles of scientific research. But, do you know what is meant by a valid and reliable data? If you still have no idea, I think this material would be perfect for you..

When we are formulating a measurement scale, i.e. the questions in the questionnaire, we have to evaluate whether the scale has been followed accurately in accordance with the principles of scientific research. In general, there are three conditions in which the questions we use are accurate: valid, has good reliability, and the can generalize to other situation. At this time, we will discuss the first criterion in research evaluation that is validity.

Practically, validity is there in every phase in research process, starting from planning process even to reporting. Before we input and process the data, we need to check the validity of instrument we used and the sample we took. The test of instrument validity includes evaluation on research samples, i.e. whether the selected sample has been valid or can represent the population closely, and evaluating on measurement tools used in research. We can conduct a pre-test to check whether our instrument (both sample and questions) are in the level of validity that can be tolerant for scientific research. While, evaluation on data validity is conducted before we input the data. We will need any statistical calculations such as t-value, standard deviation, etc. You can use statistical software like SPSS, LISREL, STATA and so on.

Definition of Validity

Data said to be valid if it is free from systematic bias and random error. Researchers define bias as the tendency of the facts obtained from respondents (or data) is influenced by other factors not included in the defined variables. Random error means the data are influenced by other factors, but in a random pattern, does not interest in one particular direction. In a simple validity definition, valid is a condition where data gained from the sample can accurately describe the real condition of measured population. Perfect validity means there is no significant difference between the mean value of the sample with mean value of a population parameter (XO = XT, XR = 0, XS = 0). And, valid argument can be said that argument is based on real facts.

Generally, there are three types of validity in research: content validity, criterion validity, and construct validity.

Content Validity

When we examine the content validity, actually what we do is evaluating whether the measurement tools that we use (in this case the research questions) are quite comprehensive covering all the dimensions of the variables studied. For example, we want to examine the store image by using questionnaire, the dimension that needs to be studied include the quality of the goods sold, variety and completeness of goods, store atmosphere, and so on. The more complete the dimensions we use, the content validity is getting better.

Criterion Validity

Criterion validity describes the extent to which a measurement tools that we use in research work according to our expectations in measuring the relationship between variables being investigated as a meaningful criterion. Variable criteria typically includes demographic and psychographic characteristics, measurement of attitudes and behaviors, or values obtained from the scale. Based on the period, the criteria validity consists of concurrent validity and predictive validity.

Concurrent validity is assessed when the data on the scale being evaluated and on the criterion variables are collected at the same time. To assess concurent validity, we may build two version of instrument, the original and the short version, then we administer both of them simultaneously to a group of respondents and the results compared. In other side, predictive validity is assessed by collecting data on the scale at one point in time and data on the criterion variables at a future time. In predictive validity the data obtained from the measurement could be used to predict a future event. For example, attitude toward instant noodle brands could be used to predict future buying behavior of instant noodle by members of certain population.

Construct Validity

Construct validity (construct validity) refers to the question: What constructs are exactly being studied or measured? When measuring the construct validity, researchers try to answer theoretical questions why the scale is used and what conclusions can be made by considering the existing theories. Construct validity is the most difficult evaluation in research because we need enough qualified theories.

Construct validity includes convergent validity, discriminant validity, and nomological validity. Convergent validity measures the extent to which the scale was positively correlated with other measures of the same construct. Convergent validity was sought to confirm the relationship between the construct in accordance with the underlying theory. Discriminant validity evaluates the extent to which measurements are not correlated with other constructs from which it should be different. And, nomological validity is the extent to which the scale correlates in theoretically predicted ways with measures of different but related constructs. A theoretical model is formulated that leads to further deductions, tests, and inferences. This nomological net should be able to interrelate systematically those constructs.

Internal Validity and External Validity

Internal validity refers to the extent to which the observed independent variables can really affect on the dependent variables. Internal validity rule says that we must use variables that are most likely to be the main cause on dependent variables, or the main cause of certain event could happen, so that other unpredicted external factors can significantly affect the dependent variables. However, in fact, internal validity is often influenced by several factors, such as: the incident between pre-test and post-test, the age of respondents, sampling methods, identification of independent and dependent variables are still vague, and the measurement tools.

Instead of internal validity refers to the ability of the manipulated independent variables can actually affect the dependent variables, the external validity aims to see whether the cause-and-effect relationships that we found in the research can be generalized to other situation, or it means that the findings can be applied even if the conditions change, both environmental conditions and the sample used. In doing research there are three forms of validation. Validation definition is a procedure to check the degree of validity particural data, sample or research instrument.

1. Fieldwork validation. It is an attempt to examine whether the fieldworker has done his job properly. Usually, supervisor contacted 10 to 25 percent of respondents to check if the interviewer has finished the interview well, how the quality of the interview, what is respondents' opinion about the interview process, and so on.


2. Sample validation. It is to check or validate whether the sample we choose are appropriate for research. The sample should have similar characteristic with the population being observed.


3. Cross validation. If we use regression analysis method we will face this type of validation. Cross validation is a validity test to examine whether the regression model continues to hold on comparable data not used in the estimation.

Validity is an absolute criterion to evaluate research findings. The invalid findings obviously cannot be used to develop a correct conclusion because they don't represent the true fact. To obtain valid findings, we need to test the validity starting from making measurements to research, select sample, process the data, to create reports.

Complete Example of Experimental Research

Have you ever been asked by your boss or maybe your professor to conduct a market test and examine the behavior of some consumer segments? Particular research, such as a study to test respondents respond toward a very new product that they have no experience about it, usually cannot be done only through questionnaires. To be able to give the right answer as we need, the respondents need to be in the situation. This is where the design of experimental research is needed.

In this material I would like to apply a case study about market testing for a new product to help you easily understand the use of experimental research. Actually, there are three types of market testing according to its purpose. First, test the market to predict sales of new products will be launched. Before you put a major investment to produce a new product, and to reduce the risk of product will be rejected by the market, the marketers need to test it first through an experiment. Second, to test the best combination of marketing mix. For example, marketers want to change the concept of product packaging with a more colorful package and the concept of the ads in magazines, they try to figure out consumers respond to the new product by conducting an experimental research. Third, experiments can also be used to find the shortcomings of existing products. In this case, the experimental research is taking part as exploratory research design.

Sample Case of Market Testing

Let's take an example, there is a big instant noodles producers in Indonesia named Mie Enak, is intending to launch its new fried noodles variety that is fried noodles with rendang flavor complete with green chili sauce. Before doing mass production, the marketing team is assigned by the Director to test whether the prototype product can get a positive attitude from the consumers (or it will be loved) and how far the new product can motivate the consumers to buy it.

Research objective: To see whether the prototype of rendang flavor fried noodle will be favored by consumers and to see whether the color of the packaging will affect consumer intention to buy the noodles.

From these goals, we are trying to arrange the elements in the experiment as below:

1. Dependent variable          = Preference and intention to buy instant noodles rendang flavor by consumers


2. Independent variable      = Product and package of the noodles.


3. Controlling variable        = The level of spicy flavor and the color of package.


4. Experiment design           = To test how the taste will affect on consumers preferences we will use a blind test method, while for packaging test will use several colors such as green, red and white colors those are arranged in a rack as in the market. And, we make a dummy magazine to see how these colors give varied effect on the consumers preferences. One respondent is only tested once.


5. Sampling method        = Simple random sampling.

Based on these elements, then we arrange some scenarios as you can see below. Scenario A and B aimed to see the effect of flavor to the intention of consumers to buy instant noodles. Given rendang identically with the spicy and savory flavors so that the quality of product variable in this experiment is the spicy flavors, which are presented in three variations: the spicy level 1, level 2, and level 3. While to look at the effect of variable color packaging we can make the simulation as in scenario C and D. Note that each respondent in this experiment only had one treatment.

Scenario A

Respondents were divided into two groups, 30 males and 30 females. The respondents presented a ready-made noodles using the same colored plates, but each plate has different taste. For example, spicy level 1, spicy level 2, and spicy level 3. Respondents were not told that the noodles taste of rendang. Respondents were asked to give an assessment on each plate from a score of 1 to 10. The greater the score given means the respondents are more favor with the noodles. Then, we look at the average value of each category of respondents as in the table and graph below.

Based on the table 1 above we can see, on average, respondents prefer spicy flavors level 2 compared to the other levels. From the two groups we also can see that the group of women tend to prefer more spicy flavor (seen from the graph that shows a V-shaped).

Scenario B

Just as in the first scenario in which respondents are presented three plates of instant noodles with different level of spicy, in this second scenario before the noodles are tasted by the respondents, they were told that the noodles are in rendang flavor. Notifying the respondent before the treatment will recall their memories of rendang so that we can know how they perceive the taste of rendang and we see their expectations for rendang noodles itself.

Based on the table above there is a change in respondents preference in noodle taste as they have different perceive what rendang should be. As we can see there is an increase in the average value of the Level 3 spicy from the first scenario. In the second scenario apparently the highest value is still in Level 2. The average value of the second segment of the respondents is 7.25, increase 0.15 points from the values obtained in the first scenario. Based on these results we can conclude that the previous experience affects consumer preferences.

Scenario C

To see which colors can draw highest attention from respondents, in this scenario we create a prototype noodles in the complete package form. After a pre-study we decided to examine three colors: white, dark green, and red colors. We listed the noodles image and its flavor in the front of the package. Then, we make the selves as in supermarkets and arrange them parallel.

From scenarios using settings such as in department stores, the results obtained can be seen in the above tables and graphs. We can see that each color has different result. The respondents tend to be more interested in the color of green packaging. Indeed, in the Indonesian market are not many manufacturers who use green package. It makes green more stand out than other colors.

Scenario D

In the fourth scenario we created two dummy magazines where three pages advertising the noodles offered. The first dummy magazine uses bright color for the background of the three noodles advertisement. While, the second dummy magazine uses dark color for the background. The three advertisement on both magazines are distributed in the center magazine (between several content pages). The advertising pages contain images visible from the front of the packaging, but the product is made of different brands by using the name that has not been used by any noodle manufacturer, for example: Tasty Noodle, Noodle Scrumptious, and Savory Noodles. After looking at the magazine, the respondent was then asked to answer the following questions:

•    Do you see any instant noodle advertisement in the magazine?

•    What color is the instant noodle's package?

•    If you saw more than one advertisement, which one do you think is the most tempting?

After answering all these questions, then the respondent was given the second magazines. Next, the respondents are asked to answer the following questions:

•    Do you see any instant noodle advertisement in the magazine?

•    What color is the instant noodle's package?

•    If you saw more than one advertisement, which one do you think is the most tempting?

•    Compared to the advertisement in the first magazine, which one is the most tempting?

Based on table 4 and graph 4 we can see that the color can determine consumer preference and intention to purchase. The green color is preferred by respondents as compared to the red and white colors. For background using bright colors, red being the most attracting advertisement, followed by the green color in the second position. However, the color green itself is superior when using dark colored backgrounds. On the average, the dark background is preferred by consumers. From these data we can conclude that the green package with a dark background pages will attract more attention of consumers.

Cara Membuat Rumusan Masalah

Dalam sebuah penelitian kita perlu membuat rumusan masalah. Rumusan masalah adalah pertanyaan-pertanyaan yang mendasari mengapa kita perlu melakukan riset. Sebelum kita memulai membahas bagaimana membuat rumusan masalah, saya asumsikan Anda sudah memilih topik untuk riset/ penelitian. Karena rumusan masalah merupakan “turunan” dari topik.

Definisi Rumusan Masalah

Berbeda dengan riset yang dilakukan oleh akademisi, tujuan riset yang dilakukan oleh sebuah entitas bisnis biasanya adalah untuk mencari peluang baru (yang belum ditemukan) dan mungkin memecahkan masalah manajerial. Dengan kata lain, riset yang dilakukan oleh entitas bisnis sebenarnya untuk membantu mereka meningkatkan kemampuan kompetisinya (competitive advantage). Sebelum memulai untuk melakukan sebuah riset seorang periset harus membuat proposal riset (research plan) terlebih dahulu, dimana langkah paling awal dan paling penting dalam menyusun proposal riset adalah membuat rumusan masalah riset. Pengertian rumusan masalah adalah pertanyaan-pertanyaan yang mendasari mengapa kita perlu melakukan riset.

Mengapa Membuat Rumusan Masalah itu Penting?

Ada dua alasan mengapa merumuskan masalah  penting sebelum memulai riset yaitu:

1. Rumusan masalah merupakan rambu-rambu mengenai apa yang dibutuhkan perusahaan dari riset sehingga hasil riset nantinya bisa memberikan solusi bagi perusahaan.
2. Definisi  masalah yang jelas membantu kita menyusun desain riset yang tepat sehingga sumber daya yang digunakan menjadi efisien. Hal ini sangat penting mengingat pelaksanaan riset seringkali membutuhkan biaya yang besar.

Meskipun membuat definisi masalah yang jelas merupakan hal yang penting dalam riset, sayangnya, seringkali pelaku dalam dunia bisnis tidak mengetahui secara jelas tujuan dari kegiatan riset yang mereka lakukan. Akibatnya, semua upaya dan sumber daya yang dikeluarkan menjadi tidak efektif dan tidak efisien karena data yang dikumpulkan ternyata memberikan solusi bagi perusahaan.  Pada pembahasan ini saya akan memberikan sejumlah tips bagaimana membuat definisi masalah yang baik dalam riset. Selamat membaca!

Bagaimana Membuat Rumusan Masalah?

Pada bagian ini kita akan belajar bagaimana membuat rumusan masalah dan untuk mempermudah Anda, saya juga menyertakan contoh rumusan masalah.

Secara umum ada lima tahapan dalam membuat rumusan masalah riset :

1. Mengumpulkan informasi yang berkaitan dengan bidang usaha atau yang memiliki pengaruh (langsung dan tidak langsung) terhadap perusahaan. Kita dapat memperoleh informasi tersebut dengan cara mengikuti perkembangan berita dari media massa, melakukan pra-riset kualitatif dengan konsumen, dan melakukan interview kepada para ahli. Dengan mengikuti perkembangan informasi terkini, kita dapat mengetahui masalah yang sedang terjadi dan kemungkinan masalah yang akan dihadapi oleh perusahaan.
2. Melakukan diskusi dengan para pembuat kebijakan di perusahaan untuk memperoleh topik riset / topik penelitian. Dalam bisnis, tujuan riset adalah untuk meningkatkan daya saing perusahaan atau mecari solusi bagi masalah perusahaan. Oleh karena topik penelitian tidak akan terlepas dari hal tersebut.
3. Setelah memperoleh topik untuk riset dari hasil diskusi dengan para pembuat kebijakan, kemudian kita pilih aspek topik yang akan diteliti. Kita perlu menyempitkan topik tersebut, tidak semua aspek topik harus diteliti karena keterbatasan sumber daya. Misalnya, topik yang ingin kita bahas adalah mengenai “Dampak Kebijakan Pemerintah terhadap Perusahaan”. Dari topik yang masih umum itu kita spesifikan lagi menjadi “Dampak Kenaikan Upah Minimum Provinsi terhadap Kinerja Perusahaan”.
4. Setelah kita menentukan topik yang akan diteliti, kita membuat broad statement (kalimat pertanyaan utama) untuk definisi masalah. Kalimat definisi masalah tidak boleh terlalu luas dan tidak boleh terlalu sempit. Menurut Malhotra, definisi masalah yang baik memiliki 3 kriteria berikut:

• Dapat membangun strategi perusahaan
• Membantu meningkatkan kemampuan kompetisi perusahaan
• Membantu meningkatkan image perusahaan.
  Contoh: Pada tahun akhir 2012 pemerintah DKI Jakarta menetapkan Upah Minimum Regional daerah Jakarta sebesar Rp 2.200.000 yang sebelumnya sebesar Rp 1.800.000.

    Permasalahan yang dihadapi oleh manajemen perusahaan: Akibat Penetapan Upah Minimum Provinsi Jakarta terhadap Kinerja Perusahaan.

    Kemudian broad statement dari rencana riset : Bagaimana Pengaruh Kenaikan Upah Minimum Provinsi terhadap Profitabilitas Perusahaan?

     5. Membuat komponen spesifik (spesific component) yang berisi pertanyaan-pertanyaan spesifik yang ingin diteliti.

Contoh rumusan masalah yang spesifik :

Dari broad statement pada contoh di atas kita dapat menyusun beberapa pertanyaan yang lebih spesifik, seperti:

• Bagaimana dampak kenaikan upah minimum provinsi tersebut terhadap efisiensi perusahaan?
• Bagaimana dampak kenaikan upah minimum tersebut terhadap profit yang dihasilkan?
• Bagaimana dampak kenaikan upah minumum terhadap kemampuan perusahaan untuk melunasi hutang?
• Bagaimana dampak kenaikan upah minimum terhadap kebijakan kepegawaian perusahaan? Apakah akan dilakukan pengurangan pegawai atau tidak?

Semantic Differential Scale and Stapel Scale

Definition of Semantic Differential Scale

Semantic differential scale is a scale consisting of 7-point measurements to measure the rank of an object based on certain attributes. Semantic differential scale has two poles that is two attributes those are mutually exclusive. Unlike the Likert scale, which is equipped with a full statement about something and ask the respondent agreement, the semantic differential scale only provides the name of object and a range with two poles opposite in nature. Then, respondents are asked to rate "how much" tendencies leads to one of the poles. Semantic differential scale is widely used to create a profile of an object and to measure the objects' image. If we want to create object's profile, then we can total the value per-nature from all respondents, next we calculate the average value that will be best attribute inherit to that object in general. If the semantic differential scale used to measure image of a topic or an object, image profiles obtained can be compared with the image of another object.

Take an example, we want to know the respondents' opinions regarding the Pepsi, a giant brand of softdrink in Indonesia.

From the example above, we figure out that respondent put Pepsi in the 3rd point. It means that in respondent thought Pepsi tends to be fairly easy to obtain. In addition, respondents also considered Pepsi tends to characterize the adults and the price is affordable. The points on the above questions illustrate attributes that describe object the most. In designing a differential semantic scale the important thing is being consistent. If at first we use a scale of 1-to-7 point, then the next question is also using the same scale. However, if from the beginning we've been using a scale -3 to +3, the next attributes are also using scale -3 to +3. If you use positive attribute in the left side, the rest attributes should be the same.

Para peneliti umumnya juga menggunakan skala semantic differential ini untuk membandingkan beberapa objek. Skala semantic differential membantu mendeskripsikan semua perbedaan dan persamaan sifat diantara objek-objek tersebut. Untuk membandingkan dua objek, peneliti bisa membandingkan nilai rerata-nya atau nilai total per sifat dan secara umum. Misalnya, untuk mengukur image sejumlah merek pakaian, peneliti dapat membandingkan nilai rerata dari masing-masing merek untuk setiap kategori sifat. Namun, ada beberapa kontroversi mengenai apakah data dari skala semantic differential dapat diperlakukan sebagai data numerik atau tidak. Jika menjadi data numerik maka peneliti dapat melakukan perhitungan (kuantifikasi) yang lebih kompleks, misalnya menggabungkan dengan regresi atau metode analisis lain.

Researchers generally use semantic differential scale to compare multiple objects' attributes.This scale helps describe all the differences and similarities in attributes of observed objects. To compare two objects, the researchers could compare the mean value or the total value per attribute and in general. For example, to measure the amount of clothing brand image, researchers can compare the average value of each brand in each category properties. However, there is some controversy as to whether the data from semantic differential scale can be treated as numeric data or not. If a numeric data, the researcher can perform calculations (quantification) are more complex, such as regression or combining with other analytical methods

Cara Membuat Skala Semantic Differential

Jika Anda tertarik untuk menggunakan skala semantic differential ini, berikut beberapa langkah menyusun skala semantic differential :

1. Tentukan dimensi-dimensi atribut yang berasosiasi dengan objek.  Misalnya kita ingin mengetahui preferensi konsumen terhadap dua provider telekomunikasi, TLKM dan INDS. Dimensi yang dapat digunakan untuk mengukur preferensi konsumen seperti: luas jangkauan jaringan, stabilitas jaringan, kelengkapan layanan, tarif, dan iklan. Dimensi tersebut haruslah merupakan hal-hal yang memang digunakan oleh konsumen secara umum untuk menilai provider telekomunikasi. Jika kita tidak dapat menemukan dimensi apa saja yang menjadi kriteria untuk mengevaluasi suatu objek maka kita perlu melakukan studi pendahuluan.


2. Dari dimensi-dimensi tersebut kita dapat menurunkan sifat-sifat objek yang diteliti. Misalnya untuk jaringan, maka sifat yang berhubungan adalah : “kuat dan lemah”, “luas dan terbatas”, “stabil dan tidak stabil”, dan sebagainya. Untuk menyusun skala semantic differential kita membutuhkan beberapa pasang sifat yang saling bertolak-belakang.


3. Sertakan petunjuk untuk menjawab sebelum pertanyaan. Petunjuk ini harus jelas agar responden dapat memberikan jawaban seperti yang kita harapkan.


4. Skala semantic differential harus konsisten. Jika sifat yang berkonotasi positif berada di kiri maka selanjutnya juga berapa di kiri.


5. Hindari menggunakan lebih dari 20 item sifat untuk menggambarkan satu objek.


6. Jika kita ingin membandingkan sifat beberapa objek, maka kita harus menggunakan skala yang sama untuk semua objek tersebut. Perbandingan dapat dilihat dari nilai rerata maupun nilai total.
   
   
   Pengertian Skala Stapel
   Nama skala ini diambil dari Jan Staple yang merupakan penemunya. Skala Stapel adalah sebuah skala dengan satu kutub terdiri dari 10 poin yaitu dari -5 hingga +5 tanpa nilai netral (nol). Pada skala ini responden diminta untuk mengidentifikasi sedekat apa objek dengan sifat yang dimaksudkan. Untuk lebih jelasnya Anda dapat melihat contoh skala Stapel berikut ini:
   
   
   Instruksi
   
   
   Pada bagian ini Anda diminta untuk mengevaluasi setiap kata yang mewakili Betamart swalayan. Anda dapat memilih satu angka dari -5 sampai +5. Semakin positif nilainya maka sifat tersebut Anda anggap semakin menggambarkan Betamart. Sebaliknya, nilai yang semakin negatif maka semakin tidak menggambarkan Betamart.

Data yang diperoleh dari skala Stapel ini dapat dianalisis dengan cara yang sama seperti pada skala semantic differential. Skala Stapel ini memiliki keunggulan diantaranya: skala Stapel tidak memerlukan studi pendahuluan seperti skala semantic differential dimana kita perlu melakukan riset pendahuluan untuk mengetahui atribut objek, dan skala Stapel dapat ditanyakan melalui telepon. Meskipun demikian, menurut beberapa peneliti aplikasi skala ini membingungkan.

Baik skala Likert, semantic differential, maupun skala Stapel ada enam hal yang perlu diputuskan ketika membuat skala tersebut:

1.    Jumlah Kategori atau Poin Skala.

Dalam membuat skala Likert, semantic, maupun Stapel kita tentukan terlebih dahulu berapa poin yang akan ada dalam skala. Apakah kita menggunakan sistem 5-poin atau 7-poin? Atau apakah kita akan menggunakan bilangan positif semua atau campuran?

Rentang poin yang panjang dapat digunakan jika responden sangat mengenal objek yang diukur. Namun, peneliti di Indonesia cenderung lebih suka menggunakan skala 5-poin karena menurut mereka pilihan yang terlalu banyak dapat membuat bingung responden dan mengurangi minat responden untuk menjawab.

2.    Seimbang versus Tidak Seimbang

Maksud seimbang di sini adalah jumlah kategori yang “baik” sama dengan yang “buruk”. Kriteria digunakan untuk skala semantic differential. Kategori tidak seimbang dapat digunakan jika jawaban dapat dimiringkan (skewed).

3.    Pilihan ganjil atau genap

Pilihan jawaban ganjil seperti skala dengan 5-poin atau 7-poin memungkinkan adanya opsi “netral” atau tidak memihak ke salah satu ekstrim. Namun, beberapa peneliti menghindari opsi ini karena “netral” dianggap bukan jawaban, atau responden tidak memiliki kecenderungan ke suatu sisi jawaban.

4.    Deskripsi verbal

Skala tidak hanya menggunakan angka-angka atau poin, namun juga dapat menggunakan gambar maupun kalimat. Pemilihan ini tergantung pada ketertarikan peneliti, responden, dan topik. Jika responden yang kita gunakan adalah siswa SD maka kita dapat menggunakan skala dalam bentuk gambar.

Contoh skala yang menggunakan gambar:

Berikan pendapat kamu mengenai film Avatar yang baru saja kamu saksikan.

Sekarang kita sudah mengetahui bentuk-bentuk inovasi skala penelitian. Yang perlu kita ingat dalam membuat skala adalah konsistensi skala yang digunakan dan gunakan instruksi yang mudah dimengerti.

How to Design a Survey Questionnaire

Questionnaire is a set of formal questions which are used to obtain information from the respondents. The questions used in a questionnaire could be written or verbal. In making a questionnaire there are three important rules we need to know. First, all questions used in questionnaire must be able to help us gaining information we need. Second, the questionnaire should be able to motivate our respondents to answer the questions. Third, a good questionnaire should be able to minimize any potential error. Therefore, it is very crucial to use an easy understood languange for our questionnaires. As wording questions in questionnaire design process will decide the level of validity of our questionnaires, thus I will give you some tips to make good questions for your questionnaire in the next session of this course.

Questionnaire Design Process

Commonly there are 7 steps in questionnaire design process which are:

1.    Specify the information needed.

In this step we need to ensure that the information obtained fully address all the components of the problem. List all problems and all possible questions. In this step you also must already have a clear idea of the target population.

2.    Specify the type of questionnaire administering method.

We have discussed about survey methods before. You can choose one or two of them.

3.    Determine the content of individual questions.

From the list of questions before, you should choose some that are really necessary for you research. Do not use double-barreled questions.

4.    Design the questions to overcome the respondent’s inability and unwillingness to answer

Before you start questioning them,you should first give them information about your research i.e. your general identity, the goals that you are trying to achieve from this questionnaire, how to answer the questions, etc. You also need to make an agreement with your respondents. Second, avoid errors of omission, telescoping, and creation. Third, put more attention to the efford required by your respondents to answer, and do not deny any sensitive questions. For sensitive questions you may need special treatment as shown bellow:

-    Place sensitive topics at the end of the questionnaire.

-    Preface the question with a statement that the behavior of interest is common.

-    Ask the question using the third-person technique.

-    Hide the question in a group of other questions that respondents are willing to answer.

-    Provide response categories rather than asking for specific figures.

-    Use randomized techniques, if appropriate.

5.    Decide on the question structure.

When we talk about question structure, we will talk about whether we use open-ended question or close-ended question. If we intend to explore all information from our respondents, we can use open-ended question structure i.e. a type of questions that let respondents give answer without limitation. Here are some examples of open-ended questions:

What is your opinion about brand X?

What do you know about Collaborating Forest Management Program?

What is your expectation from this CSR program?

Close-ended question is a type question structure that ask respondents to choose one or some answer choices given. This type of question structure is apropriate for quantitative research. There are four types of close-ended question as stated below:

1.    Simple-dichotomy question. It is a question that only provide two answer options, such as “yes” or “no” answer, but sometimes there is option “do not know”.

Example    : Have you ever heard about brand X?

a.  Yes

b.  No

2.    Determinant-choice question. It is type of question that requires respondents to choose only one from multiple choices given.

Example    :

What’s become your main consideration in buying a notebook?

a.    Functional specification

b.    Appearance

c.    Price

d.    After purchase services

e.    Other :........

Both simple dichotomy and determinant choices have same weakness that is it may let you miss any important answer which is not covered in choices we give. To overcome that problem, we can put an option “other: .......” as shown above. Respondents are asked to fill the blanks by their own words.

3.    Frequency-determination question. This is a type of question that questioning about frequency of doing something.

Example     :

How often do you update your Twitter?

a.    Once a day

b.    2 – 3 times a day

c.    4 – 5 times a day

d.    More than five times a day

4.    Scale. We can also adopt questions from research scaling, such as: Likert scale, semantic differential, staple scale, etc. You could learn them again at measurement and scaling: part 2.

Example     :

Do you intend to buy a new notebook with brand X within the next six months?

Definitely will    Probably will    Undecided    Probably    Definitely will buy    not buy        not but                will buy

1           2               3               4            5

5.    Checklist question. This type of questions that let respondents to choose more that one choices that describe them the most.

Example    :

Which product categories of XYZ cosmetic that you use in this last three months? You could choose more than one answer.

o    Compact powder

o    Moisturizer

o    Face cleanser

o    Lipstick

o    Mascara

o    Eyeliner

6.    Determine the question wording

As I said before, question wording is very crucial in designing questionnaire because it will determine whether our questionnaires are valid or not. You can see read tips for wording questions.

7.    Arrange questions in proper order

Here are some common rules in questions order:

-    The opening questions should be interesting, simple, and nonthreatening.

-    Qualifying questions should serve as the opening questions.

-    Basic information should be obtained first, followed by classification, and, finally, identification information.

-    Difficult, sensitive, or complex questions should be placed late in the sequence.

Persiapan Awal Survei House-to-House

Dalam melakukan riset ada tiga pendekatan yang dapat digunakan, yaitu pendekatan kualitatif, pendekatan kuantitatif, dan gabungan dari dua pendekatan tersebut.Pendekatan kuantitatif bertujuan untuk menguji hubungan dalam variabel penelitian. Banyak perusahaan menggunakan pendekatan ini untuk memperoleh angka yang pasti mengenai pangsa pasar, tingkat kepuasan pelanggan, skor relatif perilaku konsumen, dan sebagainya. Riset kuantitatif memang sering kali identik dengan survei yang menggunakan kuesioner, namun sebenarnya instrumen untuk riset kuantitatif ini tidak hanya kuesioner.

Dari cara pengisian kuesioner terdiri dari dua jenis, yaitu kuesioner yang diisi oleh responden (disebut self-administered questionnaire) dan kuesioner yang diisi oleh peneliti/ petugas lapangan. Beberapa buku menyebut kuesioner yang diisi oleh peneliti sebagai wawancara. Dari sisi pertanyaan kuesioner ini bisa menggunakan beberapa bentuk pertanyaan seperti pertanyaan dengan pilihan terbatas (closed-ended questions) dan pertanyaan dengan jawaban bebas (open-ended questions). Untuk riset kuantitatif biasanya menggunakan kuesioner dengan jawaban terbatas, namun sebagian menggunakan gabungan keduanya.

Bagaimana kuesioner tersebut bisa mencapai responden bisa dilakukan dengan dua cara, yaitu secara langsung dan tidak langsung. Metode langsung dilakukan dengan cara mengunjungi responden dan melakukan wawancara tatap muka dengan responden tersebut. Metode ini disebut dengan wawancara atau survei house-to-house. Sementara itu, cara tidak langsung dapat dilakukan dengan menggunakan surat atau telepon.