# Filters and attributes

Explains the options and defaults of the filters in the population form, how filters combine, how to set ratios and how SES differs from income group.

The **Population features** section of the population form decides who the synthetic consumers (personas) you generate will be. This page covers each filter's on-screen name, options and default, how filters work together, and the difference between SES and income group. The form's name, size, research topic and segmentation fields are covered in [Creating a population](/en/docs/populations/creating-a-population).

## Filter area

Filters are grouped into tabs in the left column of the form: **Demographics**, **Economic**, **Lifestyle**, **Employment**, **Social Media** and **Custom Feature**. In the Fashion and Ad Testing verticals, a vertical-specific tab is added to these. The number next to a tab name shows how many active selections that tab has.

The filter area stays dimmed and can't be clicked until you accept a segmentation on the **Segmentation Data** card. When the form opens, no filter is selected; every attribute is set to **All**.

Each attribute has its own card. The attribute's state is shown next to the card title:

| State | When it appears | Effect |
| --- | --- | --- |
| **All** | When no option is selected (the default) | No constraint; the attribute is generated with its natural distribution in data for Türkiye. |
| **Custom selection** | When you select at least one option | Only personas with the options you selected are generated. |
| **Custom ratios** | When you enter percentages with **Set ratios** | You decide each option's share of the population. |

Short lists appear as clickable chips; a selected chip shows a ✓ at the start. Long lists such as **City** and **News Source** open as a checklist with a **Search...** box at the top. You reset the selections in a single attribute with **Clear selection**, and every selection in the open tab with **Clear All Filters**. **Clear All Filters** does not touch the other tabs.

## How filters combine

- **Multiple options within the same attribute mean "or".** If you select **25-34** and **35-44** for **Age**, every persona falls into one of these two groups. The two groups are not split equally; they follow their natural proportion in data for Türkiye. If you want a specific share, use the **Set ratios** button.
- **Different attributes combine with "and".** If you select **Female** for **Gender** and **İzmir** for **City**, every persona is both female and lives in İzmir.
- **An attribute left empty covers everyone.** If you select no filters, the population is generated at the size you chose, following the distribution in data for Türkiye.
- **Conflicting selections don't stop generation.** When selections are incompatible, the system relaxes the relevant filter for personas it can't fit to your filters, and shows no warning on screen. Once the population is ready, check the result in the distribution charts.
- **Narrow SES selections can lower the number of personas.** The reason is explained in the "SES and income group" section below.

## Setting ratios

To give each option of an attribute a percentage share:

1. On the attribute card, click **Set ratios**. Your selections for that attribute are cleared and each option is given an equal percentage.
2. Enter a percentage in the box next to each option. The badge at the top of the card shows the total.
3. When the total is exactly 100, click **Use these ratios**. The button changes to **Ratios applied**.

If the total is not 100, a warning such as "Total is 90%. Make it exactly 100%, then confirm." appears below the card. To go back to selecting options instead of setting ratios, click **Back to selection**. You can't use a selection and ratios on the same attribute at the same time.

The **Set ratios** button is available on these attributes: **Age**, **Gender**, **Education**, **SES**, **Marital Status**, **Work Status**, **Household Size** and **Car & Home Ownership**.

Ratios are applied as probabilities when each persona is generated. As a result, the actual percentages can differ slightly from the values you entered, especially in small populations.

> [!IMPORTANT]
> While a ratio is unconfirmed, the **Create population** button stays disabled. Make sure you click **Use these ratios** on every card where you enter ratios.

When you set ratios on three or more attributes and one of them is **Work Status** or **Car & Home Ownership**, a yellow warning appears: **Birden fazla korelasyonlu dağılım aktif** (Multiple correlated distributions active). These two attributes are normally generated based on attributes such as age, gender and education. Setting ratios on all of them at once can, though rarely, produce inconsistent personas, which is why the warning recommends at least 100 personas.

## Filter list

The tables below list every attribute in the form by tab. The "Ratios" column shows whether the attribute has a **Set ratios** button.

### Demographics

| Attribute | Options | Ratios |
| --- | --- | --- |
| **Age** | 18-24, 25-34, 35-44, 45-54, 55-64, 65+ | Yes |
| **Gender** | Male, Female | Yes |
| **Education** | Illiterate; Literate, no school; Primary School; Middle School; High School Graduate; University Graduate; Postgraduate | Yes |
| **Marital Status** | Single, Married, Divorced, Widowed, Engaged | Yes |
| **City** | Türkiye's 81 provinces | No |

All personas are aged 18 or over. In the **City** list, İstanbul, Ankara and İzmir come first and the other provinces follow in alphabetical order. To find a province, type its name in the **Search...** box. You don't need to type Turkish characters: "canakkale" finds Çanakkale and "istanbul" finds İstanbul.

### Economic

| Attribute | Options | Ratios |
| --- | --- | --- |
| **SES** | A, B, C1, C2, D, E | Yes |
| **Credit Card Ownership** | Yes, No | No |
| **Bank Account** | Yes, No | No |
| **E-commerce Purchase** | Yes, No | No |

### Lifestyle

| Attribute | Options | Ratios |
| --- | --- | --- |
| **Alcohol Consumption** | Yes, No | No |
| **Smoking** | Yes, No | No |
| **Household Size** | 1, 2, 3, 4, 5, 6, 6+ | Yes |
| **Car & Home Ownership** | Both car and home, Car only, Home only, Neither | Yes |

In the **Smoking** filter, **Yes** covers regular smokers, people who smoke only in social settings and e-cigarette users. In the **Household Size** filter, **6** and **6+** select the same group, that is, households of six or more people.

> [!CAUTION]
> The **Alcohol Consumption** and **News Source** filters are not currently applied reliably. These two values are reassigned after the persona's lifestyle class is set, and your selection is not kept at that step. Even if you make a selection on these attributes, a large share of the personas is generated with options you did not select.

### Employment

| Attribute | Options | Ratios |
| --- | --- | --- |
| **Work Status** | Full-time employed; Part-time employed; Self-employed; Unemployed, seeking work; Unemployed, not seeking; Student; Retired; Homemaker | Yes |

### Social Media

| Attribute | Options | Ratios |
| --- | --- | --- |
| **News Source** | Doesn't watch news on TV, Doesn't watch TV, Doesn't follow news and 15 channels: A Haber, ATV, CNN Türk, Fox TV, Haber Türk, Halk TV, Kanal 7, Kanal D, KRT, NTV, Show TV, Star TV, TRT, SZC TV, TV100 | No |
| **Facebook**, **Twitter**, **WhatsApp**, **YouTube**, **Instagram**, **TikTok** | Yes, No for each | No |

In the platform filters, **Yes** means the persona uses that platform.

## SES and income group

**SES (socio-economic status)** summarises a persona's economic and social position in six classes: A, B, C1, C2, D and E. A is the highest class and E the lowest. SES is not generated directly but calculated: the persona's income group, education level and work status are scored together, and a class is assigned based on the total score.

**Income Group** measures income only and has ten bands: A1, A2, B1, B2, C1, C2, D1, D2, E1, E2. A1 is the highest decile and E2 the lowest.

The two measures are not interchangeable. SES A does not mean income group A1 or A2:

- A persona in income group B1 with a postgraduate degree and a full-time job is in SES A.
- A retiree in the highest income group (A1) whose education ended at primary school is in SES B.

Only **SES** can be filtered in the form; SES has replaced the old income group filter. Every persona still has an income group, and it appears as a separate breakdown in reports.

Because SES is calculated after generation, when you select an SES filter the system generates more personas than you asked for and keeps only those that fall into the classes you selected. If a very narrow SES selection is combined with other narrow filters, the population may be completed with fewer personas than you asked for.

> [!NOTE]
> In some reports, SES appears as **SES** and income group as **Income Group**. In Excel files they are separate columns; make sure you pick the right column when you filter results.

## Vertical-specific tabs

Fashion and Ad Testing populations have an extra tab. The fields on this tab are not filters: instead of ticking options, you type each option's share directly as a percentage. Either leave each field empty or make its total exactly 100. For a field you leave empty, the system uses the default distribution.

### Fashion: Fashion Attributes

| Field | Options |
| --- | --- |
| **Primary Shopping Channel** | **Online Ağırlıklı** (Mostly online), **AVM Ağırlıklı** (Mostly shopping malls), **Outlet Ağırlıklı** (Mostly outlets), **Karma** (Mixed), **Pasif** (Passive) |
| **Style Confidence** | **Yüksek** (High), **Orta** (Medium), **Düşük** (Low) |

If the total is not 100, the warning **Must sum to 100%.** appears below the field.

### Ad Testing: Reklam Testi Attributes

| Field | Options |
| --- | --- |
| **Kategori İlgilenimi** (Category Involvement) | **Yüksek**, **Orta**, **Düşük** |
| **Reklama Açıklık** (Ad receptivity) | **Yüksek**, **Orta**, **Düşük** |

If the total is not 100, the warning **Values should sum to 100.** appears below the field.

On both tabs, the **Create population** button stays disabled while any field has a total other than 100.

Fashion and Ad Testing personas are also split into vertical-specific archetypes (consumer types). You can't select archetypes in the form or set ratios for them; the system assigns them itself. On the detail page of an Ad Testing population, this shows as **Archetype-based distribution**. For the differences between verticals, see [Verticals](/en/docs/research/verticals).

## Attributes you can't select in the form

Some information exists on every persona but is not offered as a filter in the form:

- **Income Group:** Replaced by the **SES** filter.
- **Consumer Behavior:** Derived from the persona's personality scores after generation, so it can't be selected in advance.
- **Lifestyle class:** The **Lifestyle** tab only holds habit and ownership information; the persona's values-based lifestyle class is assigned during generation.
- **Market Segment:** Comes from the segmentation you accepted on the **Segmentation Data** card; you can't filter by segment.
- **Archetype:** Assigned by the system in Fashion and Ad Testing.

To add an attribute that isn't in the list (e.g. "Prefers eco-friendly products") to a portion of the personas, use the **Custom Feature** tab. On this tab you write the attribute as free text and enter the percentage of personas who should have it in the **Ratio** field; the rows together can't exceed 100%. A custom feature appears as a separate breakdown in the report. For details, see [Custom features](/en/docs/populations/custom-features).

## Checking your selections

The **Population summary** panel on the right of the form lists your selections under tab headings. Attributes set to **All** are not listed. Rows look like this:

- Selected age groups are separated by commas: "25-34, 35-44".
- SES classes are joined with a dash: "SES A–B".
- Yes/No selections are written with the attribute name: "Credit Card Ownership: Yes".
- Ratios are written with the attribute name and each option's percentage; options at 0 are not shown.

At the bottom of the panel are a counter showing the total number of selections (e.g. **5 features**) and the **Create population** button.

When the population is ready, open it from the **Populations** page. Charts on its detail page, such as **Age Distribution**, **Gender Distribution** and **SES Distribution**, show the actual distribution. Use them to confirm that your filters and ratios were applied as you expected. The detail page is described in [Population library](/en/docs/populations/population-library).

## Tips

- **Don't leave out the groups you want to compare.** An attribute you filter on remains a single group in the report. Instead of selecting only **Female**, set ratios such as Female 60%, Male 40%, and you can compare the two groups in the report.
- **Keep groups large enough.** Segment charts in the Price Analysis, Feature Price Impact and Competitive Analysis reports don't show segments with fewer than 5 personas; Multiple-Choice Survey shows small segments but doesn't count a difference in a segment with fewer than 30 responses as a reliable finding. If you split a 40-persona population into many small shares, some groups may not appear in the report. How breakdowns work is explained in [Segmentation](/en/docs/results/segmentation).
- **Filters don't change the cost.** A study's credit cost depends on the number of personas and the method; for details, see [Credits](/en/docs/account/credits).
- **Don't rebuild the same population.** For similar studies, select the saved population. The personas in the two studies are then exactly the same, and you can compare the results directly.

## Related pages

- [Creating a population](/en/docs/populations/creating-a-population)
- [Custom features](/en/docs/populations/custom-features)
- [Population library](/en/docs/populations/population-library)
- [Segmentation](/en/docs/results/segmentation)
- [Verticals: General Research, Fashion, Ad Testing](/en/docs/research/verticals)